如何用正则表达式移除'-'及后续两位字符以提取采样日期？

下面提供两种代码实现方案，帮你快速处理数据并转换为datetime序列：

方案一：字符串分割处理 #

通过字符串分割提取所需的日、月、年部分，补全年份后转换格式：

import pandas as pd

# 原始数据列表
raw_dates = [
    "12-15.11.12",
    "19-22.11.12",
    "26-29.11.12",
    "03-06.12.12",
    "10-13.12.12",
    "17-20.12.12",
    "19-23.12.12",
    "27-30.12.12",
    "02-05.01.13"
]

# 处理每个日期字符串
cleaned_dates = []
for s in raw_dates:
    # 拆分出采样日、后续的月年部分
    sample_day, rest = s.split("-")
    month, year_short = rest.split(".")[1:]
    # 拼接成目标格式，补全20前缀的年份
    cleaned = f"{sample_day}.{month}.20{year_short}"
    cleaned_dates.append(cleaned)

# 转换为pandas datetime序列
datetime_series = pd.to_datetime(cleaned_dates, format="%d.%m.%Y")
print(datetime_series)

方案二：正则表达式匹配 #

用正则直接提取关键部分，一步完成字符串替换：

import pandas as pd
import re

raw_dates = [
    "12-15.11.12",
    "19-22.11.12",
    "26-29.11.12",
    "03-06.12.12",
    "10-13.12.12",
    "17-20.12.12",
    "19-23.12.12",
    "27-30.12.12",
    "02-05.01.13"
]

# 正则匹配采样日、月、年，替换为目标格式
pattern = r"(\d{2})-\d{2}\.(\d{2})\.(\d{2})"
cleaned_dates = [re.sub(pattern, r"\1.\2.20\3", s) for s in raw_dates]

# 转换为datetime序列
datetime_series = pd.to_datetime(cleaned_dates, format="%d.%m.%Y")
print(datetime_series)

输出结果示例 #

0   2012-11-12
1   2012-11-19
2   2012-11-26
3   2012-12-03
4   2012-12-10
5   2012-12-17
6   2012-12-19
7   2012-12-27
8   2013-01-02
dtype: datetime64[ns]

内容的提问来源于stack exchange，提问作者crtnnn