Covariance Scoring

by microprediction

Covariance Scoring is a Data skill for Claude Code, published by microprediction in precise.

333 stars58 forkson microprediction/preciseAdded 2026/08/12Repository updated 2026/06/24

View source on GitHub Download repo ZIP

covariance-estimationcovariance-matrixinvestment-portfolioportfolio

Install in seconds

Install Covariance Scoring

Copy Covariance Scoring into your Claude Code skills folder. Run the command in your terminal, or review the source on GitHub before installing.

terminal

npx degit https://github.com/microprediction/precise/tree/main/.claude/skills/score-covariance-estimate ~/.claude/skills/score-covariance-estimate

Requires Node.js. Downloads this skill only — not the rest of the repository — into your Claude Code skills folder.

Without Node.js

git clone https://github.com/microprediction/precise.git

Clones the whole repository, then copy the skill’s own directory into your skills folder yourself.

In this catalog

Source file: .claude/skills/score-covariance-estimate/SKILL.md in microprediction/precise
Installs to: ~/.claude/skills/score-covariance-estimate
Collection: One of 6 skills cataloged from this repository
Category: Data — 668 skills

What Covariance Scoring does

Covariance Scoring scores covariance estimates with precise's assessor panel. Use it to compare estimators out of sample or against truth, especially in high dimensions where held-out likelihood can mislead.

Covariance Scoring is cataloged under Data on DirSkills. Covariance Scoring comes from a repository tagged covariance-estimation, covariance-matrix, investment-portfolio and portfolio.

Documentation

README

Excerpt

Score / rank covariance estimates

from precise import all_assessors, assessor_from_name

for A in all_assessors():
    s = A().score(cov, X_test=X_test, true_cov=Sigma_true)   # higher = better

cov is the estimate to judge; X_test is held-out data (rows = observations); true_cov is the population covariance (only available in simulation).
Each assessor exposes needs_data and needs_truth; pass what it needs. Truth-free assessors work on real data, truth-requiring ones (e.g. FrobeniusToTruth) only in simulation.
All assessors follow higher = better, so you can rank or argmax directly.

The one rule that matters in high dimensions

This is the opening of the README. Read the full README on GitHub.

Frequently asked about Covariance Scoring

What else does microprediction publish alongside Covariance Scoring?
Covariance Scoring is one of 6 skills that DirSkills catalogs from microprediction/precise, the repository it ships in. Its siblings there include Choose Covariance Estimator, Covariance Assessment and Estimate Online Covariance. Each one is a separate skill with its own page in this directory, installs the same way Covariance Scoring does, and is maintained by microprediction in that same repository. The rest of the collection is listed on the microprediction/precise page.
How does Covariance Scoring compare to other Data skills?
Covariance Scoring ranks #610 by stars among the 668 Data skills in this catalog. The most-starred ones next to it are Benchmark Methodology, Jupyter Notebook and Solana. DirSkills ranks by the star count of the repository each skill ships in, so that order reflects how popular those repositories are rather than any review of Covariance Scoring against them. Open each page to compare what they document and how they install.