Slash Command413 repo starsupdated 5mo ago

analyze

The analyze command performs comprehensive data analysis on a specified dataset using the data-explorer subagent, supporting exploratory, statistical, predictive, or complete analysis types. Use this command when you need to understand dataset structure, generate summary statistics, identify patterns and correlations, detect anomalies, perform hypothesis testing, or generate actionable insights with detailed reports and quality assessments.

View source Repository: claude-data-analysis

Install in Claude Code

Copy

mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/liangdabiao/claude-data-analysis/HEAD/.claude/commands/analyze.md -o ~/.claude/commands/analyze.md

Then start a new Claude Code session; the slash command loads automatically.

Definition

analyze.md

# Data Analysis Command

Execute data analysis on dataset `$1` with analysis type `$2` using the data-explorer subagent.

## Context
- Dataset location: @data_storage/$1
- Analysis type: $2 (exploratory, statistical, predictive, complete)
- Current working directory: !`pwd`
- Available visualization libraries: matplotlib, seaborn, plotly
- Python data science stack: pandas, numpy, scipy

## Your Task

Use the data-explorer subagent to perform comprehensive data analysis:

### 1. Data Assessment
- Load and inspect the dataset structure
- Check data types, missing values, and duplicates
- Generate initial summary statistics
- Identify data quality issues

### 2. Statistical Analysis
- Perform descriptive statistics analysis
- Calculate correlations between variables
- Identify outliers and anomalies
- Conduct appropriate statistical tests

### 3. Pattern Discovery
- Identify trends and patterns in the data
- Discover relationships between variables
- Detect seasonal patterns or cycles
- Find clusters or segments in the data

### 4. Generate Insights
- Extract key findings from the analysis
- Identify actionable insights
- Suggest areas for deeper investigation
- Recommend visualization approaches

## Analysis Types

### Exploratory Analysis
- Basic data understanding
- Summary statistics
- Data quality assessment
- Initial pattern identification

### Statistical Analysis
- Advanced statistical testing
- Correlation and regression analysis
- Hypothesis testing
- Confidence intervals

### Predictive Analysis
- Feature importance analysis
- Predictive modeling preparation
- Variable relationships
- Model recommendation

### Complete Analysis
- All of the above plus
- Comprehensive report generation
- Visualization recommendations
- Next steps planning

## Expected Output

### Analysis Report
Create a comprehensive analysis report with:
- **Executive Summary**: Key findings in plain language
- **Data Overview**: Dataset characteristics and quality
- **Statistical Findings**: Detailed statistical analysis
- **Key Insights**: Actionable discoveries
- **Recommendations**: Next steps for deeper analysis
- **Limitations**: Data and method constraints

### File Outputs
- `analysis_reports/analysis_summary_$1.md` - Detailed analysis report
- `analysis_reports/statistical_summary_$1.csv` - Statistical summary table
- `analysis_reports/data_quality_$1.json` - Data quality assessment

## Quality Assurance
- Validate all statistical calculations
- Cross-check important findings
- Document all assumptions and limitations
- Ensure reproducible analysis

## Example Usage
```bash
/analyze user_behavior.csv exploratory
/analyze sales_data.csv statistical
/analyze customer_data.csv predictive
/analyze financial_data.csv complete
```

## Notes
- Dataset should be located in the data_storage/ directory
- Analysis results will be saved to analysis_reports/ directory
- Use Task tool to delegate to data-explorer subagent
- Consider following up with /visualize command for charts