描述统计与分组分析

复刻 Stata 的 summarize, tabstat, bysort

---

基本描述统计

import pandas as pd
import numpy as np

df = pd.DataFrame({
    'age': [25, 30, 35, 40, 45, 50, 28, 32, 38, 42],
    'income': [50000, 60000, 75000, 80000, 90000, 95000, 55000, 65000, 78000, 85000],
    'gender': ['M', 'F', 'M', 'F', 'M', 'M', 'F', 'F', 'M', 'F']
})

# 基本统计
print(df['income'].describe())

# 单个统计量
print(f"均值: {df['income'].mean()}")
print(f"中位数: {df['income'].median()}")
print(f"标准差: {df['income'].std()}")
print(f"方差: {df['income'].var()}")

---

Stata 命令对照

summarize

* Stata
summarize income age
summarize income, detail

# Pandas
df[['income', 'age']].describe()

# 详细统计
df['income'].describe(percentiles=[.1, .25, .5, .75, .9])

tabstat

* Stata
tabstat income, by(gender) stat(mean sd median p25 p75)

# Pandas
df.groupby('gender')['income'].agg([
    'mean', 'std', 'median',
    ('p25', lambda x: x.quantile(0.25)),
    ('p75', lambda x: x.quantile(0.75))
])

bysort

* Stata
bysort gender: summarize income

# Pandas
df.groupby('gender')['income'].describe()

---

分组分析

# 按单个变量分组
df.groupby('gender')['income'].agg(['count', 'mean', 'std'])

# 按多个变量分组
df.groupby(['gender', 'education'])['income'].mean()

# 多列多统计
df.groupby('gender').agg({
    'income': ['mean', 'median', 'std'],
    'age': ['mean', 'min', 'max']
})

---

实战案例

案例：完整的描述统计报告

def generate_report(df):
    print("=" * 60)
    print("数据分析报告")
    print("=" * 60)

    # 样本量
    print(f"\n样本量: {len(df)}")

    # 描述统计
    print("\n【描述统计】")
    print(df[['age', 'income']].describe())

    # 按性别分组
    print("\n【按性别统计】")
    gender_stats = df.groupby('gender').agg({
        'income': ['count', 'mean', 'std'],
        'age': 'mean'
    })
    print(gender_stats)

    # 相关系数
    print("\n【相关系数】")
    print(df[['age', 'income']].corr())

generate_report(df)

---

下一步

下一节：网页数据爬取

继续！