.. pyrost documentation master file, created by sphinx-quickstart on Fri Oct 30 16:00:08 2020. You can adapt this file completely to your liking, but it should at least contain the root `toctree` directive. pyrost ====== Python Robust Speckle Tracking (**pyrost**) is a library for wavefront metrology and sample imaging based on ptychographic speckle tracking algorithm. This project takes over Andrew Morgan's `speckle_tracking <https://github.com/andyofmelbourne/speckle-tracking>`_ project as an improved version aiming to add robustness to the optimisation algorithm in the case of the high noise present in the measured data. The library is written in Python 3 and uses a C back-end written in `Cython <https://cython.org>`_. The library is capable to perform the robust version of speckle tracking algorithm, which yields an unabberated profile of the sample and the wavefront of the lens. Also it contains a set of auxiliary data processing routines, such as bad pixel masking, defocus sweep scan, wavefront reconstruction, phase model fitting, etc. All of them are listed in :class:`pyrost.STData`. pyrost includes a framework to work with CXI files, see `pyrost.cxi_ref` for more information. Speckle Tracking simulation (st_sim) ==================================== The library also contains the Speckle Tracking Simulation (**st_sim**) package. st_sim is capable to simulate one-dimensional speckle tracking scans. st_sim employs Rayleigh-Sommerfeld convolution and Fraunhofer diffraction to propagate the wavefronts. The back-end is written in C to yield the best performance. Multislice simulation (ms_sim) ============================== The multislice simulation package (**ms_sim**) is capable to propagate the wavefront through a bulky sample by the dint of multislice beam propagation algorithm. The back-end is based on FFTW library. Python Reference ================ .. toctree:: :maxdepth: 3 :caption: Contents: install tutorials/tutorials reference/api Indices and tables ================== * :ref:`genindex` * :ref:`modindex` * :ref:`search`