TransformFloat

Note

These functions are only available if single precision support is enabled, in which case the marco SPFFT_SINGLE_PRECISION is defined in config.h.

Typedefs

typedef void *SpfftFloatTransform

Transform handle.

Functions

SpfftError spfft_float_transform_create(SpfftFloatTransform *transform, SpfftFloatGrid grid, SpfftProcessingUnitType processingUnit, SpfftTransformType transformType, int dimX, int dimY, int dimZ, int localZLength, int numLocalElements, SpfftIndexFormatType indexFormat, const int *indices)

Creates a single precision transform from a single precision grid handle.

Thread-safe if no FFTW calls are executed concurrently.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [out] transform: Handle to the transform.

  • [in] grid: Handle to the grid, with which the transform is created.

  • [in] processingUnit: The processing unit type to use. Must be either SPFFT_PU_HOST or SPFFT_PU_GPU and be supported by the grid itself.

  • [in] transformType: The transform type (complex to complex or real to complex). Can be SPFFT_TRANS_C2C or SPFFT_TRANS_R2C.

  • [in] dimX: The dimension in x. The maximum allowed depends on the grid parameters.

  • [in] dimY: The dimension in y. The maximum allowed depends on the grid parameters.

  • [in] dimZ: The dimension in z. The maximum allowed depends on the grid parameters.

  • [in] localZLength: The length in z in space domain of the local MPI rank.

  • [in] numLocalElements: The number of elements in frequency domain of the local MPI rank.

  • [in] indexFormat: The index format. Only SPFFT_INDEX_TRIPLETS currently supported.

  • [in] indices: Pointer to the frequency indices. Posive and negative indexing is supported.

SpfftError spfft_float_transform_destroy(SpfftFloatTransform transform)

Destroy a transform.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

SpfftError spfft_float_transform_clone(SpfftFloatTransform transform, SpfftFloatTransform *newTransform)

Clone a transform.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] newTransform: Independent transform with the same parameters, but with new underlying grid.

SpfftError spfft_float_transform_forward(SpfftFloatTransform transform, SpfftProcessingUnitType inputLocation, float *output, SpfftScalingType scaling)

Execute a forward transform from space domain to frequency domain.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [in] inputLocation: The processing unit, to take the input from. Can be SPFFT_PU_HOST or SPFFT_PU_GPU (if GPU is set as execution unit).

  • [out] output: Pointer to memory, where the frequency domain elements are written to. Can be located at Host or GPU memory (if GPU is set as processing unit).

  • [in] scaling: Controls scaling of output. SPFFT_NO_SCALING to disable or SPFFT_FULL_SCALING to scale by factor 1 / (dim_x() * dim_y() * dim_z()).

SpfftError spfft_float_transform_backward(SpfftFloatTransform transform, const float *input, SpfftProcessingUnitType outputLocation)

Execute a backward transform from frequency domain to space domain.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [in] input: Input data in frequency domain. Must match the indices provided at transform creation. Can be located at Host or GPU memory, if GPU is set as processing unit.

  • [in] outputLocation: The processing unit, to place the output at. Can be SPFFT_PU_HOST or SPFFT_PU_GPU (if GPU is set as execution unit).

SpfftError spfft_float_transform_get_space_domain(SpfftFloatTransform transform, SpfftProcessingUnitType dataLocation, float **data)

Provides access to the space domain data.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [in] dataLocation: The processing unit to query for the data. Can be SPFFT_PU_HOST or SPFFT_PU_GPU (if GPU is set as execution unit).

  • [out] data: Pointer to space domain data on given processing unit. Alignment is guaranteed to fulfill requirements for std::complex and C language complex types.

Exceptions
  • GenericError: SpFFT error. Can be a derived type.

  • std::exception: Error from standard library calls. Can be a derived type.

SpfftError spfft_float_transform_dim_x(SpfftFloatTransform transform, int *dimX)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] dimX: Dimension in x.

SpfftError spfft_float_transform_dim_y(SpfftFloatTransform transform, int *dimY)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] dimY: Dimension in y.

SpfftError spfft_float_transform_dim_z(SpfftFloatTransform transform, int *dimZ)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] dimZ: Dimension in z.

SpfftError spfft_float_transform_local_z_length(SpfftFloatTransform transform, int *localZLength)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] localZLength: size in z of the slice in space domain on the local MPI rank.

SpfftError spfft_float_transform_local_slice_size(SpfftFloatTransform transform, int *size)

Access a transform parameter.

Parameters
  • [in] transform: Handle to the transform.

  • [out] size: Number of elements in the space domain slice held by the local MPI rank.

SpfftError spfft_float_transform_global_size(SpfftFloatTransform transform, long long int *globalSize)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] globalSize: Global number of elements in space domain. Equals dim_x() * dim_y() * dim_z().

SpfftError spfft_float_transform_local_z_offset(SpfftFloatTransform transform, int *offset)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] offset: Offset in z of the space domain slice held by the local MPI rank.

SpfftError spfft_float_transform_num_local_elements(SpfftFloatTransform transform, int *numLocalElements)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] numLocalElements: Number of local elements in frequency domain.

SpfftError spfft_float_transform_num_global_elements(SpfftFloatTransform transform, long long int *numGlobalElements)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] numGlobalElements: Global number of elements in space domain. Equals dim_x() * dim_y()

    • dim_z().

SpfftError spfft_float_transform_device_id(SpfftFloatTransform transform, int *deviceId)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] deviceId: The GPU device id used. Returns always 0, if no GPU support is enabled.

SpfftError spfft_float_transform_num_threads(SpfftFloatTransform transform, int *numThreads)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] numThreads: The exact number of threads used by transforms created from this grid. May be less than the maximum given to the constructor. Always 1, if not compiled with OpenMP support.

SpfftError spfft_float_transform_communicator(SpfftFloatTransform transform, MPI_Comm *comm)

Access a transform parameter.

Return

Error code or SPFFT_SUCCESS.

Parameters
  • [in] transform: Handle to the transform.

  • [out] comm: The internal MPI communicator.