Skip to content

Home

Does this factor possess predictive edge?

factrix is the first Polars-native Python toolkit that picks the right statistical test for each factor type. Cross-sectional, event, common factor — each gets the tests that fit its data-generating process.

GitHub

What ships

factrix exposes a small set of functions organised by category:

  • Compute — evaluate, run_metrics
  • Screening (false discovery rate (FDR)) — bhy, partial_conjunction, bhy_hierarchical
  • Inference (no FDR) — slice_pairwise_test, slice_joint_test
  • Descriptive view — by_slice, compare
  • Introspection — list_metrics, list_estimators, suggest_config

See the API reference landing for the function-flow graph, edge conventions, and the full entry-point table with category, signature summary, and when-to-reach-for guidance.

If you want Go to
Install and run a smoke test Installation · Quickstart
Understand the three-axis design (scope / signal / metric) Concepts
Compare factrix against alphalens / qlib / peers Where factrix fits
Screen a batch of factors with Benjamini-Hochberg-Yekutieli (BHY) Batch screening
Slice any metric by regime / universe / sector Slice analysis
Look up formulas and applicability Metric applicability
Read every public symbol API reference
Browse runnable notebooks Examples
Read the internal architecture Architecture