Replace pandas column with conditon-CodePudding

I have pandas df where I need to replace column name with value from row based on condition only for some columns.

 parameter.1.value| parameter.1.name |  parameter.2.value |parameter.2.name ....
 1                | abc              |   4                |xyz
 2                | abc              |   3                |xyz
 3                | abc              |   6                |xyz

Goal

 abc  |  xyz 
 1    |   4              
 2    |   3                
 3    |   6

Note:

There are other columns as well in df, also the number and location of the parameter value-name columns pair are not always same in each df

Another example with 3 pairs

parameter.1.value| parameter.1.name |  parameter.2.value |parameter.2.name|  parameter.3.value |parameter.3.name  .... 

 1               | abc              |   4                |xyz           |2 | ttt
 2               | abc              |   3                |xyz           |8 | ttt
 3               | abc              |   6                |xyz           |7 | ttt

CodePudding user response：

You can select all even columns by positions and then set new columns names by odd values in first row of DataFrame:

df1 = df.iloc[:, ::2]
df1.columns = df.iloc[0, 1::2]
print (df1)
0  abc  xyz
0    1    4
1    2    3
2    3    6

EDIT: If need select by names:

df1 = df.filter(like='parameter.value1')
df1.columns = df.filter(like='parameter.name1').iloc[0]
print (df1)
0  abc  xyz
0    1    4
1    2    3
2    3    6

EDIT1: If is possible select value and name columns use:

df1 = df.filter(like='value')
df1.columns = df.filter(like='name').iloc[0]