data-analyzer

Original source / Raw SKILL.md

---
name: data-analyzer
description: 数据分析工具，使用 pandas 进行数据统计分析。当用户需要对数据进行统计分析、计算均值、求和、描述性统计时使用。
license: MIT
compatibility: Requires Python 3.x with pandas and numpy
metadata:
  author: skillLite
  version: "1.0"
---

# Data Analyzer Skill

使用 pandas 进行数据统计分析的技能。

## 功能特性

- 支持 JSON 格式数据输入
- 支持多种统计分析操作：
  - `describe` - 描述性统计（均值、标准差、最小值、最大值等）
  - `mean` - 计算数值列均值
  - `sum` - 计算数值列求和
  - `count` - 统计行数和列信息

## 使用示例

### 描述性统计
```json
{
  "data": "[{\"name\": \"Alice\", \"age\": 25, \"score\": 85}, {\"name\": \"Bob\", \"age\": 30, \"score\": 92}]",
  "operation": "describe"
}
```

### 计算均值
```json
{
  "data": "[{\"value\": 10}, {\"value\": 20}, {\"value\": 30}]",
  "operation": "mean"
}
```

## Runtime

```yaml
entry_point: scripts/main.py
language: python
dependencies:
  - pandas>=2.0.0
  - numpy>=1.24.0
input_schema:
  type: object
  properties:
    data:
      type: string
      description: "JSON 格式的数据，如: [{\"name\": \"Alice\", \"age\": 25}, {\"name\": \"Bob\", \"age\": 30}]"
    operation:
      type: string
      description: 分析操作类型
      enum:
        - describe
        - mean
        - sum
        - count
      default: describe
  required:
    - data
```


---

## Referenced Files

> The following files are referenced in this skill and included for context.

### scripts/main.py

```python
#!/usr/bin/env python3
"""
数据分析技能 - 使用 pandas 进行数据统计分析
依赖: pandas, numpy (通过沙箱自动安装)
"""

import json
import sys

import pandas as pd


def analyze_data(data_json: str, operation: str = "describe") -> dict:
    """
    分析数据并返回统计结果
    
    Args:
        data_json: JSON 格式的数据字符串
        operation: 分析操作类型 (describe/mean/sum/count)
    
    Returns:
        包含分析结果的字典
    """
    try:
        data = json.loads(data_json)
        df = pd.DataFrame(data)
        
        if operation == "describe":
            result = df.describe(include='all').to_dict()
        elif operation == "mean":
            numeric_cols = df.select_dtypes(include=['number'])
            result = numeric_cols.mean().to_dict()
        elif operation == "sum":
            numeric_cols = df.select_dtypes(include=['number'])
            result = numeric_cols.sum().to_dict()
        elif operation == "count":
            result = {"total_rows": len(df), "columns": df.columns.tolist()}
        else:
            return {"error": f"不支持的操作类型: {operation}"}
        
        return {
            "success": True,
            "operation": operation,
            "result": result,
            "shape": {"rows": df.shape[0], "columns": df.shape[1]}
        }
        
    except json.JSONDecodeError as e:
        return {"success": False, "error": f"JSON 解析错误: {str(e)}"}
    except Exception as e:
        return {"success": False, "error": f"分析错误: {str(e)}"}


def main():
    # Read input from stdin (framework passes JSON via stdin)
    input_data = json.loads(sys.stdin.read())
    
    data_json = input_data.get("data", "[]")
    operation = input_data.get("operation", "describe")
    
    result = analyze_data(data_json, operation)
    # 处理 NaN 值，将其转换为 null
    import math
    def clean_nan(obj):
        if isinstance(obj, dict):
            return {k: clean_nan(v) for k, v in obj.items()}
        elif isinstance(obj, list):
            return [clean_nan(v) for v in obj]
        elif isinstance(obj, float) and math.isnan(obj):
            return None
        return obj
    
    result = clean_nan(result)
    print(json.dumps(result, ensure_ascii=False, default=str))


if __name__ == "__main__":
    main()

```