如何获取pandas数据框中每行的第二大值

huangapple
go评论114阅读模式
英文:

How to get second max value of each row in a pandas dataframe

问题

这会在pandas数据框的每一行中获取最大值，并创建一个名为'max'的新列，然后从该列创建一个名为maxV的新列表。

  1. df["max"] = df1.max(axis=1)
  2. maxV = df['max'].tolist()

如何在一个名为'sec_max'的新列中获取每一行的第二大值？

英文:

This gets the maximum value for each row in a pandas dataframe in a new column named 'max' and then
creates a new list named maxV from that column.

  1. df["max"] = df1.max(axis=1)
  2. maxV = df['max'].tolist()

How can i get the second max value of each row in a new column named 'sec_max'?

答案1

得分: 1

  1. import pandas as pd
  2. import numpy as np
  4. # 获取每行的第二大值的函数
  5. def get_second_max(row):
  6.     # 使用 np.partition 高效地找到第二大的值
  7.     second_largest = np.partition(row, -2)[-2]
  8.     return second_largest
  10. # 对每行应用该函数，并将结果存储在 'sec_max' 列中
  11. df['sec_max'] = df.apply(get_second_max, axis=1)
  13. # 将 'sec_max' 列转换为名为 maxV 的新列表
  14. sec_maxV = df['sec_max'].tolist()
英文: 
  1. import pandas as pd
  2. import numpy as np
  4. # Function to get the second maximum value in each row
  5. def get_second_max(row):
  6.     # Use np.partition to efficiently find the second largest value
  7.     second_largest = np.partition(row, -2)[-2]
  8.     return second_largest
  10. # Apply the function to each row and store the result in 'sec_max' column
  11. df['sec_max'] = df.apply(get_second_max, axis=1)
  13. # Convert the 'sec_max' column to a new list named maxV
  14. sec_maxV = df['sec_max'].tolist()

答案2

得分: 0

另一种解决方案，使用 np.sort：

  1. df[['sec_max', 'maxV']] = np.sort(df, axis=1)[:, -2:]
  2. print(df)

打印：

  1.    Col1  Col2  Col3  sec_max  maxV
  2. 0     1     3     2        2     3
  3. 1     6     4     5        5     6
  4. 2     7     8     8        8     8

使用的 df：

  1.    Col1  Col2  Col3
  2. 0     1     3     2
  3. 1     6     4     5
  4. 2     7     8     8
英文:

Another solution, using np.sort:

  1. df[['sec_max', 'maxV']] = np.sort(df, axis=1)[:, -2:]
  2. print(df)

Prints:

  1.    Col1  Col2  Col3  sec_max  maxV
  2. 0     1     3     2        2     3
  3. 1     6     4     5        5     6
  4. 2     7     8     8        8     8

df used:

  1.    Col1  Col2  Col3
  2. 0     1     3     2
  3. 1     6     4     5
  4. 2     7     8     8

答案3

得分: 0

  1. import pandas as pd
  2. import numpy as np
  4. data = {"col1": [420, 380, 390],
  5.         "col2": [50, 40, 45],
  6.         "col3": [102, 60, 700]}
  8. df = pd.DataFrame(data)
  10. for i in range(0,len(df.index)):
  11.     currentRow=df.iloc[i].values
  12.     max1=currentRow.max()
  13.     max2=currentRow.min()
  14.     for j in range(0,len(currentRow)):
  15.         if(currentRow[j]>max2 and currentRow[j]<max1):
  16.             max2=currentRow[j]
  17.     print(max1,max2)
英文:

here you are:

  1. import pandas as pd
  2. import numpy as np
  4. data = {&quot;col1&quot;: [420, 380, 390], 
  5.         &quot;col2&quot;: [50, 40, 45], 
  6.         &quot;col3&quot;: [102, 60, 700]}
  8. df = pd.DataFrame(data)
  10. for i in range(0,len(df.index)):
  11.     currentRow=df.iloc[i].values
  12.     max1=currentRow.max()
  13.     max2=currentRow.min()
  14.     for j in range(0,len(currentRow)):
  15.         if(currentRow[j]&gt;max2 and currentRow[j]&lt;max1):
  16.             max2=currentRow[j]
  17.     print(max1,max2)

答案4

得分: 0

  1. df['第二大'] = df.apply(lambda row: row.nlargest(2).values[-1], axis=1)
  2. print(df)

输出：

  1.    a  b   c  第二大
  2. 0  1  3   2    2
  3. 1  2  4   1    2
  4. 2  3  5  10   5
英文:

You can also use this:

df[&#39;2nd largest&#39;] = df.apply(lambda row: row.nlargest(2).values[-1],axis=1)

Example Input:

  1. import pandas as pd
  3. df = pd.DataFrame({&#39;a&#39;:[1,2,3], &#39;b&#39;:[3,4,5], &#39;c&#39;:[2,1,10]})
  4. print (df)

df:

  1.    a  b   c
  2. 0  1  3   2
  3. 1  2  4   1
  4. 2  3  5  10

code:

  1. df[&#39;2nd largest&#39;] = df.apply(lambda row: row.nlargest(2).values[-1],axis=1)
  2. print (df)

Output:

  1.    a  b   c  2nd largest
  2. 0  1  3   2            2
  3. 1  2  4   1            2
  4. 2  3  5  10            5

huangapple
  • 本文由 发表于 2023年7月18日 03:00:05
  • 转载请务必保留本文链接：https://go.coder-hub.com/76707387.html
  • dataframe
  • list
  • pandas
  • python
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定