transpose_r2() [1/2]

template<typename Scalar , Index use_library = 1>

void syten::transpose_r2	(	ConstSpan< Scalar > const	inp,
		MutSpan< Scalar >	out,
		Index const	inp_rows,
		Index const	inp_cols,
		Conj const	conj = Conj::n()
	)

Standard entry point for rank-2 transposition of any continous array, handles used backend as necessary.

Template Parameters

Scalar	scalar type of the array
use_library	method choice selection. 1 to always use the direct implementation, 2 to always use the library implementation, 1 (default) to switch between direct and library implementation as considered to be fastest.

Parameters

inp	input span (row-major)
out	output span (row-major)
inp_rows	number of rows in input range
inp_cols	number of columns in input range
conj	whether to complex-conjugate all entries

References syten::SpanImpl::MutSpan< T >::begin(), conj(), syten::Cuda::conj_copy(), std::copy(), if_constexpr, syten::YesNo< struct _prod_conj >::n(), transpose_r2_direct(), and transpose_r2_lib().

Referenced by syten::CudaDenseTensorImpl::cuda_transpose_r2(), herm(), and transpose().

Here is the call graph for this function:

Here is the caller graph for this function: