Dython is a set of Data analysis tools in pYTHON 3.x, which can let you get more insights about your data.
This library was designed with analysis usage in mind - meaning ease-of-use, functionality and readability are the core values of this library. Production-grade performance, on the other hand, were not considered.
Here are some cool things you can do with it:
Given a dataset, Dython will automatically find which features are categorical and which are numerical, compute a relevant measure of association between each and every feature, and plot it all as an easy-to-read heat-map. And all this is done with a single line:
from dython.nominal import associations associations(data)
Here's another thing - given a machine-learning multi-class model's predictions, you can easily display each class' ROC curve, AUC score and find the estimated-optimal thresholds - again, with a single line of code:
from dython.model_utils import metric_graph metric_graph(y_true, y_pred, metric='roc')
Dython can be installed directly using
pip install dython
Other installation options are available, see the installation page for more information.
See some usage examples of
model_utils.roc_graph on the examples page.
All examples can also be imported and executed from
Full documentation of all modues and public methods is available:
Read more about the
dython.nominaltools on The Search for Categorical Correlation
Read more about using ROC graphs on Hard ROC: Really Understanding & Properly Using ROC and AUC