Pandas dropna（）–从DataFrame中删除Null / NA值

1. Pandas DataFrame dropna（）函数 (1. Pandas DataFrame dropna() Function)

Pandas DataFrame dropna() function is used to remove rows and columns with Null/NaN values. By default, this function returns a new DataFrame and the source DataFrame remains unchanged.

Pandas DataFrame dropna（）函数用于删除具有Null / NaN值的行和列。 默认情况下，此函数返回一个新的DataFrame，而源DataFrame保持不变。

We can create null values using None, pandas.NaT, and numpy.nan variables.

我们可以使用None，pandas.NaT和numpy.nan变量创建空值。

The dropna() function syntax is:

dropna（）函数的语法为：

dropna(self, axis=0, how="any", thresh=None, subset=None, inplace=False)

• axis: possible values are {0 or ‘index’, 1 or ‘columns’}, default 0. If 0, drop rows with null values. If 1, drop columns with missing values.
  axis ：可能的值为{0或'index'，1或'columns'}，默认值为0。如果为0，则删除具有空值的行。 如果为1，则删除缺少值的列。
• how: possible values are {‘any’, ‘all’}, default ‘any’. If ‘any’, drop the row/column if any of the values is null. If ‘all’, drop the row/column if all the values are missing.
  如何 ：可能的值为{'any'，'all'}，默认值为“ any”。 如果为“ any”，则在任何值为null的情况下删除行/列。 如果为“全部”，则在所有值均缺失的情况下删除行/列。
• thresh: an int value to specify the threshold for the drop operation.
  thresh ：一个整数值，用于指定放置操作的阈值。
• subset: specifies the rows/columns to look for null values.
  子集 ：指定要查找空值的行/列。
• inplace: a boolean value. If True, the source DataFrame is changed and None is returned.
  inplace ：布尔值。 如果为True，则更改源DataFrame，并返回None。

Let’s look at some examples of using dropna() function.

让我们看一些使用dropna（）函数的示例。

2.熊猫使用所有Null / NaN / NaT值删除所有行 (2. Pandas Drop All Rows with any Null/NaN/NaT Values)

This is the default behavior of dropna() function.

这是dropna（）函数的默认行为。

import pandas as pd
import numpy as np

d1 = {'Name': ['Pankaj', 'Meghna', 'David', 'Lisa'], 'ID': [1, 2, 3, 4], 'Salary': [100, 200, np.nan, pd.NaT],
      'Role': ['CEO', None, pd.NaT, pd.NaT]}

df = pd.DataFrame(d1)

print(df)

# drop all rows with any NaN and NaT values
df1 = df.dropna()
print(df1)

Output:

输出：

Name  ID Salary Role
0  Pankaj   1    100  CEO
1  Meghna   2    200  None
2   David   3    NaN  NaT
3    Lisa   4    NaT  NaT

     Name  ID Salary Role
0  Pankaj   1    100  CEO

3.删除所有缺少任何值的列 (3. Drop All Columns with Any Missing Value)

We can pass axis=1 to drop columns with the missing values.

我们可以传递axis=1来删除缺少值的列。

df1 = df.dropna(axis=1)
print(df1)

Output:

输出：

Name  ID
0  Pankaj   1
1  Meghna   2
2   David   3
3    Lisa   4

4.仅当所有值都为空时才删除行/列 (4. Drop Row/Column Only if All the Values are Null)

import pandas as pd
import numpy as np

d1 = {'Name': ['Pankaj', 'Meghna', 'David', pd.NaT], 'ID': [1, 2, 3, pd.NaT], 'Salary': [100, 200, np.nan, pd.NaT],
      'Role': [np.nan, np.nan, pd.NaT, pd.NaT]}

df = pd.DataFrame(d1)

print(df)

df1 = df.dropna(how='all')
print(df1)

df1 = df.dropna(how='all', axis=1)
print(df1)

Output:

输出：

Name   ID Salary Role
0  Pankaj    1    100  NaT
1  Meghna    2    200  NaT
2   David    3    NaN  NaT
3     NaT  NaT    NaT  NaT

     Name ID Salary Role
0  Pankaj  1    100  NaT
1  Meghna  2    200  NaT
2   David  3    NaN  NaT

     Name   ID Salary
0  Pankaj    1    100
1  Meghna    2    200
2   David    3    NaN
3     NaT  NaT    NaT

5.超过空值的阈值时，DataFrame删除行/列 (5. DataFrame Drop Rows/Columns when the threshold of null values is crossed)

import pandas as pd
import numpy as np

d1 = {'Name': ['Pankaj', 'Meghna', 'David', pd.NaT], 'ID': [1, 2, pd.NaT, pd.NaT], 'Salary': [100, 200, np.nan, pd.NaT],
      'Role': [np.nan, np.nan, pd.NaT, pd.NaT]}

df = pd.DataFrame(d1)

print(df)

df1 = df.dropna(thresh=2)
print(df1)

Output:

输出：

Name   ID Salary Role
0  Pankaj    1    100  NaT
1  Meghna    2    200  NaT
2   David  NaT    NaN  NaT
3     NaT  NaT    NaT  NaT

     Name ID Salary Role
0  Pankaj  1    100  NaT
1  Meghna  2    200  NaT

The rows with 2 or more null values are dropped.

具有2个或更多空值的行将被删除。

6.定义标签以查找空值 (6. Define Labels to look for null values)

import pandas as pd
import numpy as np

d1 = {'Name': ['Pankaj', 'Meghna', 'David', 'Lisa'], 'ID': [1, 2, 3, pd.NaT], 'Salary': [100, 200, np.nan, pd.NaT],
      'Role': ['CEO', np.nan, pd.NaT, pd.NaT]}

df = pd.DataFrame(d1)

print(df)

df1 = df.dropna(subset=['ID'])
print(df1)

Output:

输出：

Name   ID Salary Role
0  Pankaj    1    100  CEO
1  Meghna    2    200  NaN
2   David    3    NaN  NaT
3    Lisa  NaT    NaT  NaT

     Name ID Salary Role
0  Pankaj  1    100  CEO
1  Meghna  2    200  NaN
2   David  3    NaN  NaT

We can specify the index values in the subset when dropping columns from the DataFrame.

当从DataFrame中删除列时，我们可以在子集中指定索引值。

df1 = df.dropna(subset=[1, 2], axis=1)
print(df1)

Output:

输出：

Name   ID
0  Pankaj    1
1  Meghna    2
2   David    3
3    Lisa  NaT

The ‘ID’ column is not dropped because the missing value is looked only in index 1 and 2.

因为缺少的值仅在索引1和2中查找，所以不会删除“ ID”列。

7.放行，NA不存在 (7. Dropping Rows with NA inplace)

We can pass inplace=True to change the source DataFrame itself. It’s useful when the DataFrame size is huge and we want to save some memory.

我们可以传递inplace inplace=True来更改源DataFrame本身。 当DataFrame很大并且我们想节省一些内存时，这很有用。

import pandas as pd

d1 = {'Name': ['Pankaj', 'Meghna'], 'ID': [1, 2], 'Salary': [100, pd.NaT]}

df = pd.DataFrame(d1)

print(df)

df.dropna(inplace=True)
print(df)

Output:

输出：

Name  ID  Salary
0  Pankaj   1   100.0
1  Meghna   2     NaN

     Name  ID  Salary
0  Pankaj   1   100.0

8.参考 (8. References)

• Python Pandas Module Tutorial
  Python Pandas模块教程
• Pandas Drop Duplicate Rows
  熊猫掉落重复行
• Pandas DataFrame dropna() API Doc
  熊猫DataFrame dropna（）API文档

翻译自: https://www.journaldev.com/33492/pandas-dropna-drop-null-na-values-from-dataframe