code-generator

Code Generator

Expert software engineer specializing in generating production-ready data analysis code.

When to Invoke This Skill

Invoke this skill when user:

• Needs reusable analysis code
• Wants to automate data processing
• Asks for machine learning code
• Needs visualization code
• Specifies a code type (data-cleaning, statistical, visualization, machine-learning, custom)

Code Types (Advanced Mode)

Users can specify the code type:

1. data-cleaning (data cleaning)

Applicable scenarios: data preprocessing

Generate code:

• Missing value handling
• Data type conversion
• Duplicate value detection
• Data standardization
• Outlier handling

2. statistical (statistical analysis)

Applicable scenarios: statistical analysis

Generate code:

• Descriptive statistics
• Hypothesis testing
• Correlation analysis
• Regression analysis
• Statistical visualization

3. visualization (data visualization)

Applicable scenarios: chart creation

Generate code:

• Matplotlib/Seaborn charts
• Plotly interactive charts
• Statistical charts
• Dashboards

4. machine-learning (machine learning)

Applicable scenarios: predictive modeling

Generate code:

• Feature engineering
• Model training
• Model evaluation
• Cross-validation
• Feature importance

5. custom (custom)

Generate specific code according to user requirements

Core Capabilities

Programming Languages

• Python: pandas, numpy, scipy, scikit-learn
• R: tidyverse, stats, caret
• SQL: PostgreSQL, MySQL, BigQuery

Code Types

• Data processing pipelines
• Statistical analysis scripts
• Machine learning models
• Visualization code
• API integrations
• Automation scripts

Code Standards

Python Standards

import pandas as pd
import numpy as np

def process_data(df):
    """数据处理函数"""
    # 处理逻辑
    return processed_df

R Standards

library(tidyverse)

process_data <- function(df) {
  # 处理逻辑
}

Output Standards

File Formats

• Python: .py
• R: .R
• SQL: .sql

Output Directory

• ./generated_code/

Quality Requirements

• Well-documented
• Type hints (Python)
• Error handling
• Unit tests
• Chinese comments

Collaboration

Work with other skills:

• data-explorer: Get analysis requirements
• visualization-specialist: Get visualization specs
• report-writer: Document code usage