mcpField
D
data-science

data-science

github.com/htlin222数据与数据库7852 次安装3 周前更新
代码仓库

关于 data-science

data-science 是用于 SQL 查询、数据探索、可视化、统计分析和 ETL 工作的技能，为使用 SQL 和 Python 处理数据集的分析人员提供模式示例。

主要功能

  1. 01编写 SQL 查询并使用窗口函数
  2. 02使用 pandas 和 NumPy 探索数据
  3. 03执行统计检验和相关性分析
  4. 04处理 BigQuery 分区与数组
  5. 05组织分析摘要和发现

适用场景

  1. 01按队列分析用户留存
  2. 02使用 RFM 分析对客户排序
  3. 03探索 CSV 文件中的数据
  4. 04检验不同群体之间的差异

安装

复制下面的命令，在终端中运行。

$npx skills add https://github.com/htlin222/dotfiles --skill data-science

安装后在 Claude Code 中输入/data-science确认已加载。

详细介绍

数据科学

数据分析、SQL 和洞察生成。

适用场景

  • 编写 SQL 查询
  • 数据分析与探索
  • 创建可视化图表
  • 统计分析
  • ETL 和数据管道

SQL 模式

常见查询
-- Aggregation with window functions
SELECT
    user_id,
    order_date,
    amount,
    SUM(amount) OVER (PARTITION BY user_id ORDER BY order_date) as running_total,
    ROW_NUMBER() OVER (PARTITION BY user_id ORDER BY order_date DESC) as recency_rank
FROM orders;

-- CTEs for readability
WITH monthly_stats AS (
    SELECT
        DATE_TRUNC('month', created_at) as month,
        COUNT(*) as total_orders,
        SUM(amount) as revenue
    FROM orders
    GROUP BY 1
),
growth AS (
    SELECT
        month,
        revenue,
        LAG(revenue) OVER (ORDER BY month) as prev_revenue,
        (revenue - LAG(revenue) OVER (ORDER BY month)) / NULLIF(LAG(revenue) OVER (ORDER BY month), 0) as growth_rate
    FROM monthly_stats
)
SELECT * FROM growth;
BigQuery 特定内容
-- Partitioned table query
SELECT *
FROM `project.dataset.events`
WHERE DATE(_PARTITIONTIME) BETWEEN '2024-01-01' AND '2024-01-31';

-- UNNEST for arrays
SELECT
    user_id,
    item
FROM `project.dataset.orders`,
UNNEST(items) as item;

-- Approximate counts for large data
SELECT APPROX_COUNT_DISTINCT(user_id) as unique_users
FROM `project.dataset.events`;

Python 分析

import pandas as pd
import numpy as np

# Load and explore
df = pd.read_csv('data.csv')
df.info()
df.describe()

# Clean and transform
df['date'] = pd.to_datetime(df['date'])
df = df.dropna(subset=['required_field'])
df['category'] = df['category'].fillna('Unknown')

# Aggregate
summary = df.groupby('category').agg({
    'value': ['mean', 'sum', 'count'],
    'date': ['min', 'max']
}).round(2)

# Visualize
import matplotlib.pyplot as plt
df.groupby('date')['value'].sum().plot(figsize=(12, 6))
plt.title('Daily Values')
plt.savefig('chart.png', dpi=150, bbox_inches='tight')

统计分析

from scipy import stats

# Hypothesis testing
t_stat, p_value = stats.ttest_ind(group_a, group_b)

# Correlation
correlation = df['x'].corr(df['y'])

# Regression
from sklearn.linear_model import LinearRegression
model = LinearRegression().fit(X, y)
print(f"R² = {model.score(X, y):.3f}")

输出格式

## Analysis Summary

**Question:** [What we're trying to answer]
**Data Source:** [Tables/files used]
**Date Range:** [Time period]

### Key Findings

1. [Finding with supporting metric]
2. [Finding with supporting metric]

### Visualization

[Chart description or embedded image]

### Recommendations

- [Actionable insight]

示例

输入：“分析用户留存率” **操作：**查询队列数据，计算留存率，并将趋势可视化

输入：“找出顶级客户” **操作：**编写用于 RFM 分析的 SQL，对用户进行分群，并总结发现

评论

支持 Markdown