Programmatic EDA

Overview

Programmatic EDA runs a systematic exploratory data analysis before deeper analysis begins, whenever you get a new dataset and need to know its shape and quality, or a surprising number shows up and you want to check the underlying data first.

What it covers

• Confirms the row count, dtypes, memory usage, and grain of the dataset first.
• Runs a null profile and flags columns above your quality thresholds.
• Detects outliers with IQR and z-score, and asks you to judge each one as signal or error.
• Summarizes distributions and flags correlated column pairs (|r| > 0.8) as possible redundancy.
• Fills a findings summary distilling the top 3 to 5 quality issues before you build anything on top of the data.

Part of Data Analytics Skills

This is one of 31 skills in the Data Analytics Skills library, in the Data Quality & Validation group.

Installation

1. Clone the repo once: git clone https://github.com/nimrodfisher/data-analytics-skills.git
2. Copy just this skill: cp -r data-analytics-skills/01-data-quality-validation/programmatic-eda ~/.claude/skills/ (or ~/.agents/skills/ for Codex)
3. Ask your agent something like "profile this dataset before I build on it" and it activates on its own.

See the Data Analytics Skills entry for the full library.