cypocketfft.fft

Python API

cypocketfft.fft.fft(double complex[:] in_arr, fct=None)

Perform the rfft for complex input

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • fct (float, optional) – Scaling factor to apply to the un-normalized transform. If not provided, defaults to \(1.0\)

Returns

The fft result as an array of complex128

Return type

numpy.ndarray

cypocketfft.fft.ifft(double complex[:] in_arr, fct=None)

Perform the inverse fft

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • fct (float, optional) – Scaling factor to apply to the un-normalized transform. If not provided, defaults to \(1.0 / N\)

Returns

The ifft result as an array of complex128

Return type

numpy.ndarray

cypocketfft.fft.irfft(double complex[:] in_arr, fct=None)

Perform the inverse rfft

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • fct (float, optional) – Scaling factor to apply to the un-normalized transform. If not provided, defaults to \(1.0 / N\)

Returns

The irfft result as an array of float64

Return type

numpy.ndarray

cypocketfft.fft.irfft_length(double complex[:] in_arr)

Calculate the irfft result length for the input array

This evaluates to \((N - 1) * 2\)

Parameters

in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

Returns

Result length

Return type

Py_ssize_t

cypocketfft.fft.rfft(double[:] in_arr, fct=None)

Perform the rfft for real input

Parameters

  • in_arr (double [:]) – Input array or typed-memoryview (real-valued)

  • fct (float, optional) – Scaling factor to apply to the un-normalized transform. If not provided, defaults to \(1.0\)

Returns

The rfft result as an array of complex128

Return type

numpy.ndarray

cypocketfft.fft.rfft_length(double[:] in_arr)

Calculate the rfft result length for the input array

This evaluates to \(N / 2 + 1\)

Parameters

in_arr (double [:]) – Input array or typed-memoryview (real-valued)

Returns

Result length

Return type

Py_ssize_t

C API

Helpers

size_t _rfft_length(double[:] in_arr)

Calculate the rfft result length for the input array

This evaluates to \(N / 2 + 1\)

Parameters

  • in_arr (double [:]) – Input array or typed-memoryview (real-valued)

Returns

The result length

size_t _irfft_length(complex_t[:] in_arr)

Calculate the irfft result length for the input array

This evaluates to \((N-1) * 2\)

Parameters
Returns

The result length

Real FFTs

Py_ssize_t _rfft(double[:] in_arr, complex_t[:] out_arr, double fct, bint use_cache=True)

Perform the rfft

Parameters

  • in_arr (double [:]) – Input array or typed-memoryview (real-valued)

  • out_arr (complex_t [:]) – Pre-allocated buffer of complex to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform (typically 1.0)

  • use_cache (bool) – If True, use the built-in plancache

Py_ssize_t _rfft_with_plan(rfft_plan* plan, double[:] in_arr, complex_t[:] out_arr, double fct)

Perform the rfft with an existing cypocketfft.wrapper.rfft_plan

Parameters

  • plan (rfft_plan*) – Pointer to a cypocketfft.wrapper.rfft_plan

  • in_arr (double [:]) – Input array or typed-memoryview (real-valued)

  • out_arr (complex_t [:]) – Pre-allocated buffer of complex to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform (typically 1.0)

  • use_cache (bool) – If True, use the built-in plancache

Py_ssize_t _irfft(complex_t[:] in_arr, double[:] out_arr, double fct, bint use_cache=True)

Perform the inverse rfft

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • out_arr (double [:]) – Pre-allocated buffer of double to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform (typically \(1.0 / N\))

  • use_cache (bool) – If True, use the built-in plancache

Py_ssize_t _irfft_with_plan(rfft_plan* plan, complex_t[:] in_arr, double[:] out_arr, double fct)

Perform the inverse rfft with an existing cypocketfft.wrapper.rfft_plan

Parameters

  • plan – Pointer to a cypocketfft.wrapper.rfft_plan

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • out_arr (double [:]) – Pre-allocated buffer of double to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform (typically \(1.0 / N\))

Complex FFTs

Py_ssize_t _cfft(complex_t[:] in_arr, complex_t[:] out_arr, double fct, bint use_cache=True)

Perform the complex fft

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • out_arr (complex_t [:]) – Pre-allocated buffer of complex to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform (typically 1.0)

  • use_cache (bool) – If True, use the built-in plancache

Py_ssize_t _icfft(complex_t[:] in_arr, complex_t[:] out_arr, double fct, bint use_cache=True)

Perform the inverse complex fft

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • out_arr (complex_t [:]) – Pre-allocated buffer of double complex to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform (typically \(1.0 / N\))

  • use_cache (bool) – If True, use the built-in plancache

Py_ssize_t _cfft_execute(complex_t[:] in_arr, complex_t[:] out_arr, double fct, bint is_forward=True, bint use_cache=True)

Helper function called by _cfft() and c:func:_icfft to perform the forward or backward transform

Parameters

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • out_arr (complex_t [:]) – Pre-allocated buffer of double complex to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform

  • is_forward (bool) – If True, perform the forward fft. If False, the inverse

  • use_cache (bool) – If True, use the built-in plancache

Py_ssize_t _cfft_with_plan(cfft_plan* plan, complex_t[:] in_arr, complex_t[:] out_arr, double fct, bint is_forward=True)

Perform the forward or inverse complex fft with an existing cypocketfft.wrapper.cfft_plan

Parameters

  • plan – Pointer to a cypocketfft.wrapper.cfft_plan

  • in_arr (complex_t [:]) – Input array or typed-memoryview (complex-valued)

  • out_arr (complex_t [:]) – Pre-allocated buffer of double complex to store the result

  • fct (double) – Scaling factor to apply to the un-normalized transform

  • is_forward (bool) – If True, perform the forward fft. If False, the inverse