Convert a patterned list into data frame-CodePudding

I've some list which is in weird shape as follows. list is saved in a var res

print(res)

output I got when I print this list

[        identity  VGG-Face_cosine            img
 0  ./clip2/0.png     1.110223e-16  .\clip1\0.png
 1  ./clip2/2.png     1.117194e-01  .\clip1\0.png
 2  ./clip2/6.png     1.117194e-01  .\clip1\0.png
 3  ./clip2/1.png     1.476002e-01  .\clip1\0.png,
         identity  VGG-Face_cosine            img
 0  ./clip2/1.png     4.440892e-16  .\clip1\1.png
 1  ./clip2/2.png     1.284562e-01  .\clip1\1.png
 2  ./clip2/6.png     1.284562e-01  .\clip1\1.png
 3  ./clip2/0.png     1.476002e-01  .\clip1\1.png,
         identity  VGG-Face_cosine            img
 0  ./clip2/2.png    -4.440892e-16  .\clip1\2.png
 1  ./clip2/6.png    -4.440892e-16  .\clip1\2.png
 2  ./clip2/0.png     1.117194e-01  .\clip1\2.png
 3  ./clip2/1.png     1.284562e-01  .\clip1\2.png,
         identity  VGG-Face_cosine            img
 0  ./clip2/3.PNG     3.330669e-16  .\clip1\3.PNG,
         identity  VGG-Face_cosine            img
 0  ./clip2/4.PNG              0.0  .\clip1\4.PNG
 1  ./clip2/8.PNG              0.0  .\clip1\4.PNG,
         identity  VGG-Face_cosine            img
 0  ./clip2/5.jpg    -2.220446e-16  .\clip1\5.jpg    ]

I've checked type of output

type(res)

output

list

my aim is to convert this into pandas df. By reading documentation what I've tried is

import pandas as pd
df_new = pd.DataFrame(res)

output I got

0   identity VGG-Face_cosine img 0 ./clip2/0.png 1.110223e-16 .\clip1\0.png 1 ./clip2/2.png 1.117194e-01 .\clip1\0.png 2 ./clip2/6.png 1.117194e-01 .\clip1\0.png 3 ./clip2/1.png 1.476002e-01 .\clip1\0.png
1   identity VGG-Face_cosine img 0 ./clip2/1.png 4.440892e-16 .\clip1\1.png 1 ./clip2/2.png 1.284562e-01 .\clip1\1.png 2 ./clip2/6.png 1.284562e-01 .\clip1\1.png 3 ./clip2/0.png 1.476002e-01 .\clip1\1.png
2   identity VGG-Face_cosine img 0 ./clip2/2.png -4.440892e-16 .\clip1\2.png 1 ./clip2/6.png -4.440892e-16 .\clip1\2.png 2 ./clip2/0.png 1.117194e-01 .\clip1\2.png 3 ./clip2/1.png 1.284562e-01 .\clip1\2.png
3   identity VGG-Face_cosine img 0 ./clip2/3.PNG 3.330669e-16 .\clip1\3.PNG
4   identity VGG-Face_cosine img 0 ./clip2/4.PNG 0.0 .\clip1\4.PNG 1 ./clip2/8.PNG 0.0 .\clip1\4.PNG
5   identity VGG-Face_cosine img 0 ./clip2/5.jpg -2.220446e-16 .\clip1\5.jpg
6   identity VGG-Face_cosine img 0 ./clip2/2.png -4.440892e-16 .\clip1\6.png 1 ./clip2/6.png 

              ........................................

output I'm expecting

        identity  VGG-Face_cosine            img
 0  ./clip2/0.png     1.110223e-16  .\clip1\0.png
 1  ./clip2/2.png     1.117194e-01  .\clip1\0.png
 2  ./clip2/6.png     1.117194e-01  .\clip1\0.png
 3  ./clip2/1.png     1.476002e-01  .\clip1\0.png
         identity  VGG-Face_cosine            img
 0  ./clip2/1.png     4.440892e-16  .\clip1\1.png
 1  ./clip2/2.png     1.284562e-01  .\clip1\1.png
 2  ./clip2/6.png     1.284562e-01  .\clip1\1.png
 3  ./clip2/0.png     1.476002e-01  .\clip1\1.png
         identity  VGG-Face_cosine            img
 0  ./clip2/2.png    -4.440892e-16  .\clip1\2.png
 1  ./clip2/6.png    -4.440892e-16  .\clip1\2.png
 2  ./clip2/0.png     1.117194e-01  .\clip1\2.png
 3  ./clip2/1.png     1.284562e-01  .\clip1\2.png
         identity  VGG-Face_cosine            img
 0  ./clip2/3.PNG     3.330669e-16  .\clip1\3.PNG
         identity  VGG-Face_cosine            img
 0  ./clip2/4.PNG              0.0  .\clip1\4.PNG
 1  ./clip2/8.PNG              0.0  .\clip1\4.PNG
         identity  VGG-Face_cosine            img
 0  ./clip2/5.jpg    -2.220446e-16  .\clip1\5.jpg

Am i missing anything here?.

CodePudding user response：

IIUC use concat - seems you get list of DataFrames:

df = pd.concat(res)

Difference of your expected ouput is only one header.