Data Analysis
data-analysis
Data cleaning, exploratory analysis, statistical testing, and visualization report generation for decision-ready insights.
Install
clawhub install data-analysis
Cross-tool: Claude Code ~/.claude/skills/ · Qoder ~/.qoder/skills/ · TRAE ~/.trae/skills/
When to Use
Trigger this skill when performing exploratory data analysis, statistical testing, or generating data insight reports.
Core Workflow
Data Cleaning
- Identify missing value patterns: MCAR vs MAR vs MNAR, handle accordingly
- Detect outliers with IQR or Z-score, flag rather than delete
- Standardize data types: date formats, numeric precision, string encoding
Exploratory Analysis
- Start with distributions (histograms, box plots), then relationships (scatter, correlation matrix)
- Group comparisons: group statistics + confidence intervals, not just means
- Time series: trend, seasonality, anomalies
Statistical Testing
- Two groups: t-test (normal) or Mann-Whitney U (non-normal)
- Multiple groups: ANOVA + post-hoc (Tukey HSD)
- Correlation: Pearson (linear) or Spearman (monotonic), report p-value and effect size
Visualization
- Chart selection: line for trends, bar for comparison, histogram for distribution, scatter for relationships
- Annotate key data points, outliers, and threshold lines
- Each chart accompanied by a one-sentence insight