v0.1.0
Fast, Polars-native data profiling with interactive HTML reports.
Added
ProfileReport— 3-line data profiling for Polars and Pandas DataFrames- 5 column analyzers: numeric, categorical, boolean, datetime, text
- Dataset overview: row/column counts, missing totals, duplicate rows, memory usage, type distribution
- 7 data quality alerts: HIGH_MISSING, CONSTANT, HIGH_CARDINALITY, DUPLICATES, HIGH_ZEROS, SKEWED, IMBALANCED
- Pearson correlation heatmap (Polars-native)
- Interactive HTML report with ECharts (histogram, bar, pie, word cloud, heatmap)
- Programmatic access:
to_dict(),to_json(),.stats,.alerts,.correlations - Two profiling modes:
"complete"(deep analysis) and"overview"(fast summary) - Pandas auto-conversion via
pl.from_pandas() - CSV and Parquet file path ingestion
- PEP 561
py.typedmarker
Full Changelog: https://github.com/dataxid/dataxid-profiling/releases/tag/v0.1.0