Automatic EDA

v0.4.0

One-call exploratory profiling: overview, per-column statistics, correlation structure and data-quality alerts — with a ViewX-ready report payload.

Overview

profile() inspects a DataFrame and returns an EDAReport: types, missingness, uniqueness, numeric moments, categorical top values, the Pearson correlation matrix and a list of alerts that flag suspicious columns before they poison a model.

API

profile

profile(data, columns=None) -> EDAReport

EDAReport

EDAReport.overview

overview: n_rows, n_columns, n_numeric, n_categorical, n_datetime, n_duplicate_rows, memory_kb

EDAReport.columns

columns -> DataFrame

EDAReport.alerts

alerts -> list of {severity, column, issue, detail}

EDAReport.to_report_data

to_report_data(include_figures=False, data=None) -> dict

dict

Exporters

to_dataframe / to_dict / to_markdown / to_json

rep.to_dataframe() · rep.to_dict() · rep.to_markdown() · rep.to_json()