eda
Exploratory Data Analysis for tabular data. Use this skill when analyzing value distributions, checking for missing data, computing correlations, examining class balance, or generating data quality reports.
📖 中文导读:Exploratory Data Analysis for tabular data. Use this skill when analyzing value distributions, checking for missing data, computing correlations, examining class balance, or generating data quality reports. 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。
name: eda description: Exploratory Data Analysis for tabular data. Use this skill when analyzing value distributions, checking for missing data, computing correlations, examining class balance, or generating data quality reports.
Exploratory Data Analysis (EDA)
Analyze tabular datasets to understand distributions, data quality, and relationships between variables.
When to Use
- Understanding a new dataset before modeling
- Checking data quality (missing values, outliers, duplicates)
- Analyzing target variable distribution for classification/regression
- Identifying correlations between features
- Generating summary statistics
Available Tasks
| Task | Command | Description |
|---|---|---|
| Column Distribution | eda-column-dist |
Analyze value distribution for a specific column |
Task Documentation
Detailed task templates are available in tasks/:
tasks/column_distribution.md- Full documentation for column distribution analysis
Quick Start
# Analyze distribution of a column
eda-column-dist --source <path> --column <name>
# Save report to file
eda-column-dist --source <path> --column <name> --output report.md
Output Format
All EDA scripts produce markdown reports with:
- Task metadata (source, column, timestamp)
- Summary statistics
- Distribution tables or visualizations (as text)
- Observations and potential issues
Best Practices
- Start with data-connector - Verify data access and schema before EDA
- Check target variable first - Understand class balance for classification tasks
- Look for missing patterns - Missing data may not be random (MCAR/MAR/MNAR)
- Document findings - Save reports for reproducibility
Future Tasks (Planned)
- Missing data analysis
- Correlation matrix
- Outlier detection
- Duplicate detection
- Target class balance
- Full EDA report (combines all tasks)