PyMLUtils

Reusable ML utilities I kept rewriting across projects — so I packaged them properly.

Install

pip install pymlutils

What's Included

• preprocessing — Smart imputers, encoders, scalers with fit/transform API
• evaluation — Classification/regression report generators with visualisations
• explainability — SHAP wrappers, feature importance plots
• pipelines — Pre-built sklearn pipeline templates for common use cases
• validation — Time-series aware cross-validators

Quick Example

from pymlutils.evaluation import ClassificationReport
from pymlutils.preprocessing import SmartImputer
from sklearn.ensemble import RandomForestClassifier

imputer = SmartImputer(strategy='auto')
X_clean = imputer.fit_transform(X_train)

clf = RandomForestClassifier()
clf.fit(X_clean, y_train)

report = ClassificationReport(clf, X_test, y_test)
report.show()
report.plot_roc()
report.plot_confusion()

Why This Exists

I was writing the same preprocessing boilerplate and evaluation helpers on every project. Packaged them up, put them on GitHub. More people found it than I expected.

Built by Shebin S Illikkal — Shebinsillikkal@gmail.com | Kerala, India