Maximum Likelihood Phaser (MaxLP) for holographic, or reference-enhanced single particle imaging (SPI) experiments. For the description of the experiment and reconstruction algorithm, see the following papers:

• K Ayyer, "Reference-enhanced x-ray single-particle imaging." Optica 7.6 (2020): 593-601
• A Mall & K Ayyer, "Holographic single particle imaging for weakly scattering, heterogeneous nanoscale objects." arXiv:2210.10611

This is a pure python 3 package, which has the following dependencies:

• cupy, which in turn needs CUDA
• numpy
• h5py
• mpi4py (for multiple GPUs)

To perform a quick simulation, run the following commands:

$ mkdir -p data/test/
$ ./make_data.py -c config_sim.ini
$ ./recon.py -c config_sim.ini 1

The output can be found in data/test/output_001.h5

If you have access to multiple GPUs, you can parallelize the reconstruction using MPI. Before one can run the reconstruction, you have to create a 'devices' file which tells the program which GPU number each rank must run on. This is a simple text file with one number per line, referring to the GPU ID as seen by nvidia-smi.

$ mpirun -np 4 python recon.py -d devices.txt 10

In order to see benefits from MPI, you should increase the number of frames and/or the number of rotational samples.