Simulations for Reference-enchanced Single Particle Imaging

This code contains code for simulating reconstructions from variable-reference holographic data. The algorithm and simulation details are described in the following paper:

Reference-enhanced X-ray Single Particle Imaging, Optica 7(6), 593-601 (2020)

The program is pure-Python and runs on a CUDA-enabled GPU using the CuPy library. Additionally, the following libraries are used

• numpy for CPU math
• h5py for data storage
• scipy for image filtration in Shrinkwrap
• mpi4py to run the reconstruction on multiple GPUs

Typically cupy and mpi4py require custom installations depending upon system versions of CUDA and MPI respectively. Other packages can be installed easily using PyPi or Anaconda.

A default configuration file has been provided which has the same parameters as those described in the paper, along with an explanation of the various options.

Quick start

To generate data similar to the paper, just run

./make_data.py

You can use the -m option to generate just the real-space object or the -d option to just generate the photon data from the object already generated.

The reconstruction can be run after generating the data by running

./emc.py 100

where the number refers to the number of iterations to run.

The output is saved in Numpy-format .npy files in the data/ directory. For advanced options, run either command with the -h option.