Pandas小技巧

• Column重命名

df.columns = ['a', 'b', 'c']

• union

df = pd.concat([df1, df2])

• Index去重，包含column df3 = df3[~df3.index.duplicated(keep='first')]

• Columns去重

df = pd.concat([df1, df2, df2]).drop_duplicates(subset=['col1'], keep=False)

• 遍历每一行

for index, row in df.iterrows():
    # row是dict形式，key: column name
    print(row['col1'])
for row in df.itertuples():
    # row是tuple形式，第一个元素是index
    print(row[1])