Note: Clicking on the option with a triangle bullet ▸ next to it will expand/collapse its detailed information.; Type name written in PascalCase typically refer to a NEML2 object type, oftentimes a primitive tensor type.; The 🔗 symbol means that the tensor value can be cross-reference another object. See Model parameters (revisited) for details.; You can always use Ctrl+F or Cmd+F to search the entire page.

Available objects and their input file syntax

Refer to System Documentation for detailed explanation about this system.

CSVSR2

Construct a SR2 from a CSV file. Each component of the SR2 should be provided as a column in the CSV. By default the entire CSV is read in order of the columns. A subset/different order of columns can be selected using column_indices or column_names. Each row of the CSV corresponds to a batch. The default batch shape is the number of rows in the CSV.

column_indices Indices of CSV columns.

Type: list of non-negative number

column_names Names of CSV columns.

Type: list of string

csv_file Path to the CSV file

Type: string

delimiter Delimiter used to parse the CSV file. Options are SPACE, TAB, SEMICOLON, COMMA

Type: EnumSelection
Default: COMMA

no_header Whether the CSV file has a header row.

Type: bool
Default: false

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

starting_row Starting row of the CSV file (0-indexed). Rows before this row are ignored. This should be the header row if the CSV file has a header, otherwise it should be the first row of the data. By default the starting row is the 0th row.

Type: number

Detailed documentation link

CSVScalar

Construct a Scalar from a CSV file. Each component of the Scalar should be provided as a column in the CSV. By default the entire CSV is read in order of the columns. A subset/different order of columns can be selected using column_indices or column_names. Each row of the CSV corresponds to a batch. The default batch shape is the number of rows in the CSV.

column_indices Indices of CSV columns.

Type: list of non-negative number

column_names Names of CSV columns.

Type: list of string

csv_file Path to the CSV file

Type: string

delimiter Delimiter used to parse the CSV file. Options are SPACE, TAB, SEMICOLON, COMMA

Type: EnumSelection
Default: COMMA

no_header Whether the CSV file has a header row.

Type: bool
Default: false

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

starting_row Starting row of the CSV file (0-indexed). Rows before this row are ignored. This should be the header row if the CSV file has a header, otherwise it should be the first row of the data. By default the starting row is the 0th row.

Type: number

Detailed documentation link

CSVVec

Construct a Vec from a CSV file. Each component of the Vec should be provided as a column in the CSV. By default the entire CSV is read in order of the columns. A subset/different order of columns can be selected using column_indices or column_names. Each row of the CSV corresponds to a batch. The default batch shape is the number of rows in the CSV.

column_indices Indices of CSV columns.

Type: list of non-negative number

column_names Names of CSV columns.

Type: list of string

csv_file Path to the CSV file

Type: string

delimiter Delimiter used to parse the CSV file. Options are SPACE, TAB, SEMICOLON, COMMA

Type: EnumSelection
Default: COMMA

no_header Whether the CSV file has a header row.

Type: bool
Default: false

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

starting_row Starting row of the CSV file (0-indexed). Rows before this row are ignored. This should be the header row if the CSV file has a header, otherwise it should be the first row of the data. By default the starting row is the 0th row.

Type: number

Detailed documentation link

CenterMillerIndex

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: MillerIndex 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterQuaternion

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: Quaternion 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterR2

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: R2 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterR3

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: R3 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: R4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterRot

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: Rot 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterSFFR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: SFFR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterSFR3

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: SFR3 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterSR2

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: SR2 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterSSR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: SSR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterSWR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: SWR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterScalar

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: Scalar 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterTensor

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: Tensor 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterVec

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: Vec 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterWFFR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: WFFR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterWR2

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: WR2 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterWSR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: WSR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

CenterWWR4

Compute interval centers along an intermediate dimension.

dim Intermediate dimension to compute centers

Type: number

points The input tensor to be centered

Type: WWR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceMillerIndex

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: MillerIndex 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceQuaternion

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: Quaternion 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceR2

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: R2 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceR3

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: R3 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: R4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceRot

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: Rot 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceSFFR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: SFFR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceSFR3

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: SFR3 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceSR2

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: SR2 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceSSR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: SSR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceSWR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: SWR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceScalar

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: Scalar 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceTensor

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: Tensor 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceVec

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: Vec 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceWFFR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: WFFR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceWR2

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: WR2 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceWSR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: WSR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

DifferenceWWR4

Compute finite differences along an intermediate dimension.

dim Intermediate dimension to take the finite difference

Type: number

points The input tensor to be differenced

Type: WWR4 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

FillR2

Construct a R2 with a vector of Scalars. The vector length must be 1, 3, 6, or 9. When vector length is 1, the Scalar value is used to fill the diagonals; when vector length is 3, the Scalar values are used to fill the respective diagonal entries; when vector length is 6, the Scalar values are used to fill the tensor following the Voigt notation; when vector length is 9, the Scalar values are used to fill the tensor in the row-major fashion.

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Scalars used to fill the R2

Type: list of Scalar> 🔗

Detailed documentation link

FillRot

Construct a Rot from a vector of Scalars.

method Fill method, options are 'modified' and 'standard'.

Type: string
Default: modified

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Scalars used to fill the Rot

Type: list of Scalar> 🔗

Detailed documentation link

FillSR2

Construct a SR2 with a vector of Scalars. The vector length must be 1, 3, or 6. See the full documentation on neml2::SR2 on the dispatch of fill method. When vector length is 1, the Scalar value is used to fill the diagonals; when vector length is 3, the Scalar values are used to fill the respective diagonal entries; when vector length is 6, the Scalar values are used to fill the tensor following the Voigt notation.

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Scalars used to fill the SR2

Type: list of Scalar> 🔗

Detailed documentation link

FillWR2

Construct a WR2 from a vector of Scalars.

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Scalars used to fill the WR2

Type: list of Scalar> 🔗

Detailed documentation link

FullMillerIndex

Construct a full MillerIndex with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullQuaternion

Construct a full Quaternion with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullR2

Construct a full R2 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullR3

Construct a full R3 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullR4

Construct a full R4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullRot

Construct a full Rot with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullSFFR4

Construct a full SFFR4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullSFR3

Construct a full SFR3 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullSR2

Construct a full SR2 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullSSR4

Construct a full SSR4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullSWR4

Construct a full SWR4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullScalar

Construct a full Scalar with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullTensor

Construct a full Tensor with given batch shape. Tensor values are set to the specified value.

base_shape Base shape

Type: tensor shape

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullVec

Construct a full Vec with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullWFFR4

Construct a full WFFR4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullWR2

Construct a full WR2 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullWSR4

Construct a full WSR4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

FullWWR4

Construct a full WWR4 with given batch shape. Tensor values are set to the specified value.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

value Value to fill the tensor with

Type: number

Detailed documentation link

GaussianScalar

Construct a Scalar with values sampled from a Gaussian profile at given points. The Gaussian is \(g(x) = h \exp(-\tfrac{1}{2} z^2)\) with \(z = (x - c) / w\), where \(h\) is the height, \(w\) is the width, and \(c\) is the center.

center The Gaussian center in the domain

Type: Scalar 🔗

height The Gaussian height

Type: Scalar 🔗

points The coordinates to evaluate the Gaussian at

Type: Scalar 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

width The Gaussian width

Type: Scalar 🔗

Detailed documentation link

GaussianTensor

Construct a Tensor with values sampled from a Gaussian profile at given points. The Gaussian is \(g(x) = h \exp(-\tfrac{1}{2} z^2)\) with \(z = (x - c) / w\), where \(h\) is the height, \(w\) is the width, and \(c\) is the center.

center The Gaussian center in the domain

Type: Scalar 🔗

height The Gaussian height

Type: Scalar 🔗

points The coordinates to evaluate the Gaussian at

Type: Tensor 🔗

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

width The Gaussian width

Type: Scalar 🔗

Detailed documentation link

IdentityTensor

Construct an identity Tensor with given batch shape.

batch_shape Batch shape

Type: tensor shape

n Diagonal size of the identity tensor, i.e., base shape of the identity tensor will be (n,n)

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

LinspaceMillerIndex

Construct a MillerIndex linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: MillerIndex 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: MillerIndex 🔗

Detailed documentation link

LinspaceQuaternion

Construct a Quaternion linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Quaternion 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Quaternion 🔗

Detailed documentation link

LinspaceR2

Construct a R2 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: R2 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: R2 🔗

Detailed documentation link

LinspaceR3

Construct a R3 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: R3 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: R3 🔗

Detailed documentation link

LinspaceR4

Construct a R4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: R4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: R4 🔗

Detailed documentation link

LinspaceRot

Construct a Rot linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Rot 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Rot 🔗

Detailed documentation link

LinspaceSFFR4

Construct a SFFR4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SFFR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SFFR4 🔗

Detailed documentation link

LinspaceSFR3

Construct a SFR3 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SFR3 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SFR3 🔗

Detailed documentation link

LinspaceSR2

Construct a SR2 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SR2 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SR2 🔗

Detailed documentation link

LinspaceSSR4

Construct a SSR4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SSR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SSR4 🔗

Detailed documentation link

LinspaceSWR4

Construct a SWR4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SWR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SWR4 🔗

Detailed documentation link

LinspaceScalar

Construct a Scalar linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Scalar 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Scalar 🔗

Detailed documentation link

LinspaceTensor

Construct a Tensor linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Tensor 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Tensor 🔗

Detailed documentation link

LinspaceVec

Construct a Vec linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Vec 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Vec 🔗

Detailed documentation link

LinspaceWFFR4

Construct a WFFR4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WFFR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WFFR4 🔗

Detailed documentation link

LinspaceWR2

Construct a WR2 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WR2 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WR2 🔗

Detailed documentation link

LinspaceWSR4

Construct a WSR4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WSR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WSR4 🔗

Detailed documentation link

LinspaceWWR4

Construct a WWR4 linearly spaced on the batch/intermediate dimensions. See neml2::dynamic_linspace, neml2::intmd_linspace, or neml2::base_linspace for a detailed explanation.

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WWR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WWR4 🔗

Detailed documentation link

LogspaceMillerIndex

Construct a MillerIndex logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: MillerIndex 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: MillerIndex 🔗

Detailed documentation link

LogspaceQuaternion

Construct a Quaternion logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Quaternion 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Quaternion 🔗

Detailed documentation link

LogspaceR2

Construct a R2 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: R2 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: R2 🔗

Detailed documentation link

LogspaceR3

Construct a R3 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: R3 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: R3 🔗

Detailed documentation link

LogspaceR4

Construct a R4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: R4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: R4 🔗

Detailed documentation link

LogspaceRot

Construct a Rot logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Rot 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Rot 🔗

Detailed documentation link

LogspaceSFFR4

Construct a SFFR4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SFFR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SFFR4 🔗

Detailed documentation link

LogspaceSFR3

Construct a SFR3 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SFR3 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SFR3 🔗

Detailed documentation link

LogspaceSR2

Construct a SR2 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SR2 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SR2 🔗

Detailed documentation link

LogspaceSSR4

Construct a SSR4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SSR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SSR4 🔗

Detailed documentation link

LogspaceSWR4

Construct a SWR4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: SWR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: SWR4 🔗

Detailed documentation link

LogspaceScalar

Construct a Scalar logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Scalar 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Scalar 🔗

Detailed documentation link

LogspaceTensor

Construct a Tensor logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Tensor 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Tensor 🔗

Detailed documentation link

LogspaceVec

Construct a Vec logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: Vec 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: Vec 🔗

Detailed documentation link

LogspaceWFFR4

Construct a WFFR4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WFFR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WFFR4 🔗

Detailed documentation link

LogspaceWR2

Construct a WR2 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WR2 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WR2 🔗

Detailed documentation link

LogspaceWSR4

Construct a WSR4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WSR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WSR4 🔗

Detailed documentation link

LogspaceWWR4

Construct a WWR4 logarithmically spaced on the batch dimensions. See neml2::dynamic_logspace, neml2::intmd_logspace, or neml2::base_logspace for a detailed explanation.

base The base of the logarithm

Type: number
Default: 10

dim Where to insert the new dimension

Type: number

end The ending tensor

Type: WWR4 🔗

group Dimension group to apply the operation. Options are: intermediate, dynamic

Type: EnumSelection
Default: dynamic

nstep The number of steps with even spacing along the new dimension

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

start The starting tensor

Type: WWR4 🔗

Detailed documentation link

MillerIndex

Construct a MillerIndex from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

MillerIndexFromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

MultiColumnCSVScalar

Construct a two-dimensional Scalar from a CSV file. A subset of columns can be selected using column_indices or column_names. By default, the CSV is interpreted as column-major, i.e., each column in the CSV corresponds to one row of the 2D Scalar. This behavior can be altered via the indexing option.

column_indices Indices of CSV columns.

Type: list of non-negative number

column_names Names of CSV columns.

Type: list of string

csv_file Path to the CSV file

Type: string

delimiter Delimiter used to parse the CSV file. Options are SPACE, TAB, SEMICOLON, COMMA

Type: EnumSelection
Default: COMMA

indexing Indexing interpretation. Options are ROW_MAJOR, COLUMN_MAJOR

Type: EnumSelection
Default: COLUMN_MAJOR

no_header Whether the CSV file has a header row.

Type: bool
Default: false

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

starting_row Starting row of the CSV file (0-indexed). Rows before this row are ignored. This should be the header row if the CSV file has a header, otherwise it should be the first row of the data. By default the starting row is the 0th row.

Type: number

Detailed documentation link

Orientation

An orientation, internally defined as a set of Modified Rodrigues parameters given by \( r = n \tan{\frac{\theta}{4}} \) with \( n \) the axis of rotation and \( \theta \) the rotation angle about that axis. However, this class provides a variety of ways to define the orientation in terms of other, more common representations.

angle_convention Euler angle convention, 'Kocks', 'Roe', or 'Bunge'

Type: string
Default: kocks

angle_type Type of angles, either 'degrees' or 'radians'

Type: string
Default: degrees

input_type The method used to define the angles, 'euler_angles' or 'random'

Type: string
Default: euler_angles

normalize If true do a shadow parameter replacement of the underlying MRP representation to move the inputs farther away from the singularity

Type: bool
Default: false

quantity Number (batch size) of random orientations

Type: non-negative number
Default: 1

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Input Euler angles, as a flattened n-by-3 matrix

Type: list of number

Detailed documentation link

Quaternion

Construct a Quaternion from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

QuaternionFromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

R2

Construct a R2 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

R2FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

R3

Construct a R3 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

R3FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

R4

Construct a R4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

R4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

RandomMillerIndex

Construct a random MillerIndex with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomQuaternion

Construct a random Quaternion with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomR2

Construct a random R2 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomR3

Construct a random R3 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomR4

Construct a random R4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomRot

Construct a random Rot with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomSFFR4

Construct a random SFFR4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomSFR3

Construct a random SFR3 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomSR2

Construct a random SR2 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomSSR4

Construct a random SSR4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomSWR4

Construct a random SWR4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomScalar

Construct a random Scalar with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomTensor

Construct a random Tensor with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

base_shape Base shape

Type: tensor shape

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomVec

Construct a random Vec with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomWFFR4

Construct a random WFFR4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomWR2

Construct a random WR2 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomWSR4

Construct a random WSR4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

RandomWWR4

Construct a random WWR4 with given batch shape. Tensor values are set to random values. Values are drawn uniformly from the sample space.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

max Maximum random value.

Type: number

min Minimum random value.

Type: number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

Rot

Construct a Rot from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

RotFromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

SFFR4

Construct a SFFR4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

SFFR4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

SFR3

Construct a SFR3 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

SFR3FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

SR2

Construct a SR2 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

SR2FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

SSR4

Construct a SSR4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

SSR4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

SWR4

Construct a SWR4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

SWR4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

Scalar

Construct a Scalar from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

ScalarFromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

SymmetryFromOrbifold

Returns a tensor of symmetry operations for a given symmetr group represented in orbifold notation.

orbifold A string giving the orbifold representation of the group, for example 432 for the typical cubic crystal system defined by chiral octahedral symmetry

Type: string

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

Detailed documentation link

Tensor

Construct a Tensor from a vector values. The vector will be reshaped according to the specified batch shape.

base_shape Base shape

Type: tensor shape

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

Vec

Construct a Vec from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

VecFromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

WFFR4

Construct a WFFR4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

WFFR4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

WR2

Construct a WR2 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

WR2FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

WSR4

Construct a WSR4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

WSR4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

WWR4

Construct a WWR4 from a vector values. The vector will be reshaped according to the specified batch shape.

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

values Values in this (flattened) tensor

Type: list of number

Detailed documentation link

WWR4FromTorchScript

Get the tensor from torch script. The torch script should define a Module with named_buffers that stores the tensor to load. Refer to tests/regression/liquid_infiltration/gold/generate_load_file.py for an example

batch_shape Batch shape

Type: tensor shape

intermediate_dimension Intermediate dimension

Type: non-negative number

shape_manipulation_args A list of arguments corresponding to each shape manipulation operation. The number of entries should match the number of operations in 'shape_manipulations'. Each entry is a tensor shape that encodes the arguments for the corresponding operation. For operations that do not require any argument, an empty shape, i.e. (), should be used.

Type: list of tensor shape

shape_manipulations A list of shape manipulation operations to apply to the created tensor. Supported operations are: intmd_flatten, dynamic_expand, dynamic_reshape, dynamic_squeeze, intmd_expand, dynamic_transpose, dynamic_flatten, intmd_reshape, intmd_squeeze, INVALID, intmd_unsqueeze, dynamic_movedim, intmd_transpose, dynamic_unsqueeze, intmd_movedim

Type: MultiEnumSelection
Default: (INVALID)

tensor_name Key of named_buffers to extract the tensor from.

Type: string

torch_script Name of the torch script file.

Type: string

Detailed documentation link

Table of Contents

Available objects and their input file syntax

CSVSR2

CSVScalar

CSVVec

CenterMillerIndex

CenterQuaternion

CenterR2

CenterR3

CenterR4

CenterRot

CenterSFFR4

CenterSFR3

CenterSR2

CenterSSR4

CenterSWR4

CenterScalar

CenterTensor

CenterVec

CenterWFFR4

CenterWR2

CenterWSR4

CenterWWR4

DifferenceMillerIndex

DifferenceQuaternion

DifferenceR2

DifferenceR3

DifferenceR4

DifferenceRot

DifferenceSFFR4

DifferenceSFR3

DifferenceSR2

DifferenceSSR4

DifferenceSWR4

DifferenceScalar

DifferenceTensor

DifferenceVec

DifferenceWFFR4

DifferenceWR2

DifferenceWSR4

DifferenceWWR4

FillR2

FillRot

FillSR2

FillWR2

FullMillerIndex

FullQuaternion

FullR2

FullR3

FullR4

FullRot

FullSFFR4

FullSFR3

FullSR2

FullSSR4

FullSWR4

FullScalar

FullTensor

FullVec

FullWFFR4

FullWR2

FullWSR4

FullWWR4

GaussianScalar

GaussianTensor

IdentityTensor

LinspaceMillerIndex

LinspaceQuaternion

LinspaceR2

LinspaceR3

LinspaceR4

LinspaceRot

LinspaceSFFR4

LinspaceSFR3

LinspaceSR2

LinspaceSSR4

LinspaceSWR4

LinspaceScalar

LinspaceTensor

LinspaceVec