当一列不为null时，python pandas匹配两列之间的值 [英] python pandas match values between two columns when one column is not null

问题描述

在这里找不到类似的问题.

Didn't find the similar question here.

请在下面找到表格:

         A       B       C      D
  0     pen     nan     dfds    1238
  1     Apple   pen     fsd      324
  2     Peach   nan     kd       878
  3     grape   peach   jil      9kj
  4     laptop  nan     lks      873p
  5     light   grape   kje      7623d
  6     nan     grape   3r43     kj23
  7     nan     grape   3fdf     8734d

1. 如果列B不为空，则将B中的值与A中的值进行比较，并尝试查找匹配的值.例如笔"在列A的第一行中="pen"在B列的第二行.
2. 如果识别出匹配的值，则需要在A列中找到索引.笔"是一个匹配值，"pen"的索引是A列中的值为0.

我的预期输出是:

         A       B        C      D 
  0     pen     nan     dfds    1238
  2     Peach   nan     kd       878
  3     grape   peach   jil      9kj

并保留原始索引号，如输出示例所示

And keep original index number as in the output example

我知道如何在A和B之间进行匹配.我的代码是

I know how to do the matching job between A and B. My code is

df2=df[df[['A','B']].nunique(axis=1)==1]

但是当B列不为null时，我不知道如何添加条件.而且我不想进行循环迭代，因为数据集非常大.

But i don't know how to add the condition when column B is not null. And i don't want to do loop iterations since the dataset is super large.

非常感谢！

推荐答案

我认为，在您问题的第1点中，您的意思是" C "列不为空吗?但是无论如何，我都会在列" B "中进行演示.

I think, in point 1 of your question, do you mean column "C" is not null?? But whatever, I'll demonstrate it by taking column "B".

为此，您必须创建一个仅包含非空值的新数据框.

For this, you have to create a new dataframe containing not null values only.

df_not_null = df.dropna(subset=['B'])

然后，您可以比较任何您想比较的东西.

Then you can compare whatever you want to compare.

df2 = df_not_null[df_not_null[['A','B']].nunique(axis=1)==1]

这篇关于当一列不为null时，python pandas匹配两列之间的值的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！