lspsuite is a set of Python tools (and bash scripts) related to LSP simulations. They may be useful for certain stages of creating and analyzing the simulations.

This suite incorporates code written both by Scott Feister and by Gregory Ngirmang -- the latter, as part of his "lspreader" repository: https://github.com/noobermin/lspreader

The primary use case has been personal, but I'm sharing these tools in case others find them useful.

This module requires Python 3.6+ and a working implementation of MPI (e.g. OpenMPI or MPICH). Installation requires git.

OS X users: Prior to installing dependencies, ensure an adequate Python installation by following this guide. The Python that ships with OS X may not work well with some required dependencies.

• numpy
• scipy
• matplotlib
• h5py
• mpi4py

The dependencies may be installed according to the directions on their webpages, or with any Python package manager that supports them. For example, one could use pip to install them as

pip install numpy scipy matplotlib h5py mpi4py

NOTE: If you are on a cluster where you do not have write permissions to the python installation directory, you may need to add "--user" to your pip and setup calls here and below. E.g.

pip install --user numpy scipy matplotlib h5py mpi4py

As an alternate to pip, one could also use Anaconda Python to install the dependencies

conda install numpy scipy matplotlib h5py
conda install -c conda-forge mpi4py

After installing the required packages, we may install lspsuite.

One way to install lspsuite is via

pip install git+https://github.com/phyzicist/lspsuite.git

To update lspsuite at a later date

pip install --upgrade git+https://github.com/phyzicist/lspsuite.git

An alternative way to install lspsuite is via

git clone https://github.com/phyzicist/lspsuite.git
cd lspsuite/
python setup.py install

If you installed with the "--user" flag and "~/.local/bin" is not on your system path, add it to your path (e.g. by appending the line "PATH=$PATH:$HOME/.local/bin" to "~/.bashrc"). This will allow you to use any command-line tools from this repository (if any exist yet!).

You will need to read through the source code for descriptions of the functions of this package. There is unfortunately no formal documentation.

import lspsuite as ls

filenames = ls.listp4(".", prefix="flds")
timelist = ls.times(filenames) # works only on flds and sclr

domains1, header1 = ls.read_flds("flds41.p4")
domains2, header2 = ls.read_sclr("sclr41.p4")
frames = ls.read_pmovie("pmovie5.p4")
pext = ls.read_pext("pext1.p4")
values, labels = ls.read_history("history.p4")
griddict = ls.read_grid("grid.p4")
regdict = ls.read_regions("regions.p4")
voldict = ls.read_volumes("volumes.p4")

First, make a file like this one ("mytraj.py")

import lspsuite as ls

p4dir = "/simulations/run1" # A folder containing pmovie LSP outputs with the initial positions SAK flag on
ls.mpitraj(p4dir) # Creates a file called 'traj.h5' in p4dir

Second, call the script with MPI and at least two processors.

mpirun -np 4 python mytraj.py

To uninstall lspsuite

pip uninstall lspsuite