Skip to content
Kitbay
Skill

exploratory-data-analysis

MITUpdated on Nov 9, 202656,764GitHub starsIncludes scripts

A K-Dense skill for a first, cautious look at scientific data files kept on the local machine.

What the skill teaches

A K-Dense skill for a first, cautious look at scientific data files kept on the local machine. Bundled command-line scripts, which use only the Python standard library for the core formats, produce aggregate profiles of CSV, TSV and JSON tables, and optional inspectors read NumPy arrays, HDF5 metadata, FASTA and FASTQ sequences and basic image metadata. It audits missing values and possible data leakage, checks how outliers and transformations change results, and gives a template for an EDA report. The skill treats file contents as untrusted, never prints raw rows or identifiers, makes no network calls, and refuses unknown formats.

Capabilities

  • Profile CSV, TSV and JSON tables without printing raw rows
  • Inspect NumPy, HDF5, FASTA/FASTQ and image metadata
  • Audit missing values and data leakage
  • Check outlier and transformation sensitivity
  • Scaffold an exploratory data analysis report