如何删除熊猫数据框中的唯一行

Question

index                                            SUBJECT
1                                                   test
2                                                  Hello
3                                                  Hello
4                               PRC review - phone calls

拆卸后

index                                            SUBJECT
2                                                  Hello
3                                                  Hello

我只想删除基于“ SUBJECT”列的行。 这个怎么做？

Answer 1

使用duplicated

例如：

import pandas as pd

df = pd.DataFrame({"SUBJECT": ["test", "Hello", "Hello", "PRC review - phone calls"]})
df = df[df.duplicated(subset=["SUBJECT"], keep=False)]
print(df)

输出：

  SUBJECT
1   Hello
2   Hello

Answer 2

您可以这样做：

# get count for each value
s = df.SUBJECT.value_counts()

# get only those that appear more than once
repeated = set(s[s > 1].index.values)

# filter the data-frame base
result = df[df.SUBJECT.isin(repeated)]

print(result)

输出量

   index SUBJECT
1      2   Hello
2      3   Hello

Answer 3

检查一下：

df.loc[(df.groupby('SUBJECT').count()>1).sum(axis=1),:]

Answer 4

解决方案1：

使用loc ..

>>> df.loc[df.duplicated(keep=False), :]
  SUBJECT
1   Hello
2   Hello

解决方案2：

groupby + 转换的另一种方法..

>>> df[df.groupby('SUBJECT')['SUBJECT'].transform('size') > 1]
  SUBJECT
1   Hello
2   Hello