nabu.preproc.ctf_cuda module

class nabu.preproc.ctf_cuda.CudaCTFPhaseRetrieval(shape, geo_pars, delta_beta, padded_shape='auto', padding_mode='reflect', translation_vh=None, normalize_by_mean=False, lim1=1e-05, lim2=0.2, use_rfft=True, fftw_num_threads=None, fft_num_threads=None, logger=None, cuda_options=None, fft_backend='skcuda')[source]

Bases: CTFPhaseRetrieval

Cuda back-end of CTFPhaseRetrieval

Initialize a Contrast Transfer Function phase retrieval.

Parameters:

  • geo_pars (GeoPars) – the geometry description

  • delta_beta (float) – the delta/beta ratio

  • padded_shape (str or tuple, optional) – Padded image shape, in the form (num_rows, num_columns) i.e (vertical, horizontal). By default, it is twice the image shape.

  • padding_mode (str) – Padding mode. It must be valid for the numpy.pad function

  • translation_vh (array, optional) – Shift in the form (y, x). It is used to perform a translation of the image before applying the CTF filter.

  • normalize_by_mean (bool) – Whether to divide the (padded) image with its mean before applying the CTF filter.

  • lim1 (float >0) – the regulariser strenght at low frequencies

  • lim2 (float >0) – the regulariser strenght at high frequencies

  • use_rfft (bool, optional) – Whether to use real-to-complex (R2C) FFT instead of usual complex-to-complex (C2C).

  • fftw_num_threads (bool or None or int, optional) – DEPRECATED - please use fft_num_threads instead.

  • fft_num_threads (bool or None or int, optional) – Number of threads to use for FFT. If a number is provided: number of threads to use for FFT. You can pass a negative number to use N - fft_num_threads cores.

  • logger (optional) – a logger object

set_input(data)
retrieve_phase(image, output=None)[source]

Perform padding on an image. Please see the documentation of CTFPhaseRetrieval.retrieve_phase().