Distribution Analyzer

Pro v1.0.0 1 view

Automated distribution analysis for numeric dataset variables with statistics, visualizations, and transformation recommendations.

What You Get

Automate hours of exploratory data analysis work. Get comprehensive distribution insights, normality tests, and actionable transformation recommendations in seconds.

The Problem

Data analysts and scientists spend significant time manually analyzing distributions across multiple variables before modeling. This includes calculating statistics, creating visualizations, running normality tests, and determining appropriate transformations - repetitive work that delays actual analysis.

The Solution

Run a single command to analyze all numeric variables in your dataset. The script calculates comprehensive statistics (mean, median, std, quartiles, skewness, kurtosis), classifies distribution shapes, runs appropriate normality tests (Shapiro-Wilk or Kolmogorov-Smirnov based on sample size), recommends transformations, and optionally generates a visualization grid with histograms and box plots.

How It Works

  1. 1 Load CSV or Excel file and identify all numeric columns for analysis
  2. 2 Calculate comprehensive summary statistics including central tendency, spread, and shape measures
  3. 3 Classify distribution shapes based on skewness and kurtosis thresholds
  4. 4 Run appropriate normality tests (Shapiro-Wilk for n<5000, K-S for larger samples)
  5. 5 Generate transformation recommendations based on distribution characteristics
  6. 6 Optionally create visualization grid with histograms (KDE overlay) and box plots

What You'll Need

  • CSV or Excel file with numeric data
  • Python 3.9+ (auto-handled by uv)

Get This Skill

Requires Pro subscription ($9/month)

Have Feedback?

Help us improve this skill by sharing your ideas and suggestions.

Request Improvements

Examples

Customer Demographics Dataset

Analyzing a customer dataset with age, income, and spending score columns. Demonstrates basic distribution analysis with symmetric distributions and appropriate standardization recommendations for variables with heavy tails.

View example

IoT Sensor Data with Large Sample

Analyzing sensor readings with 6000 observations to trigger Kolmogorov-Smirnov test instead of Shapiro-Wilk. Demonstrates correct test selection for large samples, handling of zero-inflated error counts, and identification of normally distributed sensor measurements alongside skewed operational metrics.

View example

Sales Data with Skewed Distributions

Analyzing sales metrics including revenue, units sold, profit margin, and discount rate. Demonstrates handling of mixed distribution types - right-skewed revenue requiring log transformation, normally distributed profit margins, and moderate skew requiring square root transformation.

View example
View all examples