README for coord-learn
Installation and Dependencies
Install CSD Software.
Python dependencies are listed in (and, if using Anaconda, installed using)
the environment.yml file.
Usage
The _harvest#.py scripts are examples of how to search the CSD for entries
containing the specified elements, and save those entries' labels as
sample sets.
CSD -> entries as .npy files ("samples")
_train04.py is an example of how to use the harvested entries to generate
data sets and train a Perceptron to identify atom coordination numbers.
This Perceptron is then evaluated on a test data set, determining the
accuracy of the Perceptron on each element.
samples -> .csv file containing the accuracies and # samples for
each element
The _plotacc**.py scripts are examples of how to plot the accuracy data
generated for each element.