darianyang/wetree

resampling tree visualization for westpa h5 files

wetree

Make a resampling tree visual after running a WE simulation.

---

This tutorial contains the code necessary to generate a resampling tree with a Wepy simulation hdf5 file. A small, example hdf5 file is provided. This file has a 25 cycle, 12 trajectory dataset and an observable used for coloring of nodes. This tree can be visualized in Gephi.

---

Adapted from: https://gitlab.com/nmroussey/resampling_tree_tutorial Converted from compatibility with wepy h5 to westpa h5 files (in progress).

---

Installation and Requirements

If you have generated a Wepy hdf5 file, your environment should already contain all necessary packages.

Wepy: pip install wepy[all]

NumPy, NetworkX, and sys are also required.

---

Contents of the Repository

1. run_restree_gen.py

The script to run to generate the tree. Arguments are: input h5 file and output file name.

2. restree/

The codes required to make the tree.

3. example_input.wepy.h5

A sample wepy output file.

Generating The Tree

In your enviornment that contains Wepy do:

python run_restree_gen.py example_input.wepy.h5 test_outfile.gexf

This will generate a .gexf file for use in 'Gephi'. Data included is nodes and edges, weights/probabilities (for node size), and an extra piece of information named COM, for coloring the tree.

NOTE: No merging and cloning is visible in this tree due to the short length of the simulation. This is not a mistake in the tree generation code.

---

References

Software Packages

'Wepy'

'NumPy'

'Gephi'

Papers

'Wepy: A Flexible Software Framework for Simulating Rare Events with Weighted Ensemble Resampling' Lotz, S., & Dickson, A., ACS Omega, 2020

---

ENV: conda install -c condo-forge mdtraj pip install wepy[all] Downgrade numpy to 1.23.* (pip install numpy=1.23)

wetree