feyn 3.3.1

Feyn is a symbolic regression package named after Richard Feynman, that uses the QLattice as a simulator to generate models.

Feyn

Quick start

Feyn is available as Python3.8+ package through pip. You can install it with the following command:

richard@feyn:~$ pip3 install feyn

Once installed, go to your preferred Python environment and follow along with this example.

Running a QLattice

If you're using the community edition of a QLattice then you can instantiate it by:

import feyn

ql = feyn.QLattice()

Auto run

The quickest way to get started is to use the auto_run function on the QLattice. First we will make a classification problem with feyn.datasets.make_classification.

from feyn.datasets import make_classification

train, test = make_classification()
models = ql.auto_run(train, output_name = 'y', kind = 'classification')

This returns a list of fitted models that are the best the QLattice has sampled, sorted by ascending loss.

Evaluate

The model with the lowest loss is models[0]. We can evaluate that model with the plot function and it's ROC curve.

best = models[0]
best.plot(train, test)
best.plot_roc_curve(test)