Core Concepts
Workflow Cheat Sheet
Quick reference mapping developer questions to Featuresmith APIs
Use this cheat sheet as a rapid lookup guide to map common data engineering tasks directly to Featuresmith SDK entrypoints and CLI commands.
SDK & CLI Task Mapping
| Task / Question | Python SDK Function | CLI Command |
|---|---|---|
| Ingest CSV, Excel, Parquet, or DataFrame | fs.load(source) | featuresmith analyze data.csv |
| Extract raw statistical profile without rules | fs.profile(dataset) | featuresmith analyze data.csv |
| Check atomic quality rules & target leakage | fs.analyze(ds, target_column="y") | featuresmith analyze data.csv --target y |
| Run complete 10-reviewer dataset code review | fs.review(ds, target_column="y") | featuresmith review data.csv --target y |
| Run diff-aware review (compare with previous snapshot) | fs.review(ds, previous=prev, target_column="y") | featuresmith review data.csv --previous prev.csv --target y |
| Extract 0–100 ML Readiness Scorecard | fs.score(review_result) | featuresmith review data.csv --target y |
| Compare snapshot versions (Dataset Diff) | fs.diff(old, new, target_column="y") | featuresmith diff old.csv new.csv --target y |
| Extract RuleFinding objects from diff | fs.diff_findings(diff_result) | featuresmith diff old.csv new.csv --format json |
| Render text report for console / logs | fs.render(review_result) | featuresmith review data.csv |
| Render diff text report for console | fs.render_diff(diff_result) | featuresmith diff old.csv new.csv |