两个pandas模糊匹配

weixin_38901465

已于 2022-12-29 20:25:11 修改

阅读量556

点赞数

文章标签： pandas python 开发语言

于 2022-12-22 20:38:19 首次发布

原文链接：https://blog.csdn.net/weixin_43734080/article/details/121228769?ops_request_misc=&request_id=&biz_id=102&utm_term=%E4%B8%A4%E4%B8%AApandas%E6%A8%A1%E7%B3%8A%E5%8C%B9%E9%85%8D&utm_medium=distribute.pc_search_result.none-task-blog-2~all~sobaiduweb~default-0

版权

import pandas as pd
import re

df_a = pd.DataFrame({
    "keyide" : ["a", "b", "c", "d"],
    "keyword" : ["numpy", "pandas", "matplotlib", "map"]
})

df_b = pd.DataFrame({"senid" : range(10,19),
                        "sentence" : [
                            "怎样用pan1das实现merge",
                            "Python之Numpy详细教程",
                            "怎么使用Pandas批量拆分与合并Excel文件？",
                            "怎样使用pandas的map和apply函数？",
                            "深度学习之tensorflow简介",
                            "tensorflow和numpy的关系",
                            "基于sklearn的一些机器学习的代码",
                            "基于fghjfj的一些机器学习的代码",
                            "落地鞥过来了干辣椒雷锋精神代理费据了解"
                        ]})

key_word_dict = df_a.to_dict(orient="dict").get("keyword")

df_b['keyids'] = df_b["sentence"].map(lambda sss: [keyid for keyid, key_word in key_word_dict.items() if re.search(key_word, sss, re.IGNORECASE)])

df_result = pd.merge(
    left = df_b.explode("keyids"),
    right = df_a,
    how='left',
    left_on = "keyids",
    right_on = df_a.index)

print(df_result)