classification-report-as-df 0.1.0

sklearn-style classification report as a pandas DataFrame

classification-report-as-df

Small, single-module utility that converts scikit-learn's text-style classification report into an easy-to-manipulate pandas DataFrame. For easier display in IPython environments (e.g. Jupyter), the report can also be returned as a pandas Styler.

Example usage

General scenario:

import numpy as np
from classification_report_as_df import classification_report_as_df

y_true = np.array([0, 1, 0, 1, 1])
y_pred = np.array([0, 0, 0, 1, 1])

df = classification_report_as_df(
    y_true=y_true,
    y_pred=y_pred
)

Jupyter notebook pretty-rendering scenario:

import numpy as np
from classification_report_as_df import classification_report_as_df

y_true = np.array([0, 1, 0, 1, 1])
y_pred = np.array([0, 0, 0, 1, 1])

classification_report_as_df(
    y_true=y_true,
    y_pred=y_pred,
    decimal_places_for_display=2
)

Development

Clone repository:

git clone https://github.com/Michal58/classification-report-as-df
cd classification-report-as-df

Install dependencies:

pip install classification_report_as_df[dev]

or (from root directory):

pip install -e '.[dev]'

You can run tests calling (from root directory):

pytest

License

This project is licensed under the MIT License — see the LICENSE file for details.