dbt_split Module Reference

Routines to split blocks and to convert between tensors with different block sizes. More...

Functions/Subroutines
subroutine, public	dbt_split_blocks_generic (tensor_in, tensor_out, blk_size_1, blk_size_2, blk_size_3, blk_size_4, nodata)
	Split tensor blocks into smaller blocks. More...
subroutine, public	dbt_split_blocks (tensor_in, tensor_out, block_sizes, nodata)
	Split tensor blocks into smaller blocks of maximum size PRODUCT(block_sizes). More...
subroutine, public	dbt_split_copyback (tensor_split_in, tensor_out, summation)
	Copy tensor with split blocks to tensor with original block sizes. More...
subroutine, public	dbt_make_compatible_blocks (tensor1, tensor2, tensor1_split, tensor2_split, order, nodata1, nodata2, move_data)
	split two tensors with same total sizes but different block sizes such that they have equal block sizes More...
subroutine, public	dbt_crop (tensor_in, tensor_out, bounds, move_data)

Detailed Description

Routines to split blocks and to convert between tensors with different block sizes.

Author

Patrick Seewald

Function/Subroutine Documentation

dbt_split_blocks_generic()

subroutine, public dbt_split::dbt_split_blocks_generic	(	type(dbt_type), intent(inout)	tensor_in,
		type(dbt_type), intent(out)	tensor_out,
		integer, dimension(:), intent(in), optional	blk_size_1,
		integer, dimension(:), intent(in), optional	blk_size_2,
		integer, dimension(:), intent(in), optional	blk_size_3,
		integer, dimension(:), intent(in), optional	blk_size_4,
		logical, intent(in), optional	nodata
	)

Split tensor blocks into smaller blocks.

Parameters

tensor_in	Input tensor
tensor_out	Output tensor (splitted blocks)
blk_size_i	block sizes for each of the tensor dimensions
nodata	don't copy data from tensor_in to tensor_out

Author

Patrick Seewald

Definition at line 68 of file dbt_split.F.

Here is the call graph for this function:

Here is the caller graph for this function:

dbt_split_blocks()

subroutine, public dbt_split::dbt_split_blocks	(	type(dbt_type), intent(inout)	tensor_in,
		type(dbt_type), intent(out)	tensor_out,
		integer, dimension(ndims_tensor(tensor_in)), intent(in)	block_sizes,
		logical, intent(in), optional	nodata
	)

Split tensor blocks into smaller blocks of maximum size PRODUCT(block_sizes).

Parameters

tensor_in	Input tensor
tensor_out	Output tensor (split blocks)
block_sizes	block sizes for each of the tensor dimensions
nodata	don't copy data from tensor_in to tensor_out

Author

Patrick Seewald

Definition at line 454 of file dbt_split.F.

Here is the call graph for this function:

dbt_split_copyback()

subroutine, public dbt_split::dbt_split_copyback	(	type(dbt_type), intent(inout)	tensor_split_in,
		type(dbt_type), intent(inout)	tensor_out,
		logical, intent(in), optional	summation
	)

Copy tensor with split blocks to tensor with original block sizes.

Parameters

tensor_split_in	tensor with smaller blocks
tensor_out	original tensor

Author

Patrick Seewald

Definition at line 526 of file dbt_split.F.

Here is the call graph for this function:

Here is the caller graph for this function:

dbt_make_compatible_blocks()

subroutine, public dbt_split::dbt_make_compatible_blocks	(	type(dbt_type), intent(inout)	tensor1,
		type(dbt_type), intent(inout)	tensor2,
		type(dbt_type), intent(out)	tensor1_split,
		type(dbt_type), intent(out)	tensor2_split,
		integer, dimension(ndims_tensor(tensor1)), intent(in), optional	order,
		logical, intent(in), optional	nodata1,
		logical, intent(in), optional	nodata2,
		logical, intent(in), optional	move_data
	)

split two tensors with same total sizes but different block sizes such that they have equal block sizes

Parameters

move_data	memory optimization: transfer data s.t. tensor1 and tensor2 may be empty on return
tensor1_split	tensor 1 with split blocks
tensor2_split	tensor 2 with split blocks
nodata1	don't copy data of tensor 1
nodata2	don't copy data of tensor 2

Definition at line 786 of file dbt_split.F.

Here is the call graph for this function:

Here is the caller graph for this function:

dbt_crop()

subroutine, public dbt_split::dbt_crop	(	type(dbt_type), intent(inout)	tensor_in,
		type(dbt_type), intent(out)	tensor_out,
		integer, dimension(2, ndims_tensor(tensor_in)), intent(in)	bounds,
		logical, intent(in), optional	move_data
	)

Author

Patrick Seewald

Definition at line 933 of file dbt_split.F.

Here is the call graph for this function:

Here is the caller graph for this function: