导入期间跳过行

Question

i know how to skip rows... and how to handle different header, but how can I manage different headers?

In my example, I have a CSV file like this:

Header_row1; Header_row2;....
2 ;3 ;...
2 ;3 ;...
2 ;3 ;...
2 ;3 ;...
Header_row1; Header_row2;....
2 ;3 ;...
2 ;3 ;...
2 ;3 ;...
Header_row1; Header_row2;....
2 ;3 ;...
2 ;3 ;...

I want a panda dataframe with only one header for my further calculations.

Thanks a lot.

Edit: After a few comments: A part of my code:

for h in range(len(dpath)):
  path = lidar_save + dpath[h]

  #Combine seperate files to one file over the periode.
  data_month = pd.DataFrame()
  data_month_std = pd.DataFrame()
  wind_rec_gz = glob.glob(path+'/*.csv')
  print('Read: ', wind_rec_gz[0])

  df = pd.read_csv(wind_rec_gz, header=0, sep=';',encoding = 'unicode_escape')

Answer 1

I'm not sure there's a way to do so while importing. But you can do that after import:

df = pd.read_csv('your_csv_file')

# this check every row if they are different from the column names
s = df.ne(df.columns, axis=1).any(axis=1)

# s is
#0      True
#1      True
#2      True
#3      True
#4     False
#5      True
#6      True
#7      True
#8     False
#9      True
#10     True
#dtype: bool


# keep only those rows
df = df[s]

Output:

   Header_row1  Header_row2 ....
0           2            3   ...
1           2            3   ...
2           2            3   ...
3           2            3   ...
5           2            3   ...
6           2            3   ...
7           2            3   ...
9           2            3   ...
10          2            3   ...

Answer 2

The variable wind_rec_gz is a list ['/media/..../rge/merge_2019-04-04.csv'] (look at the brackets). Get rid of the brackets and it should work (along with Quang Hoang's answer to get rid of the extra header rows).

wind_rec_gz = r'/media/..../rge/merge_2019-04-04.csv'

df = pd.read_csv(wind_rec_gz, sep=';', header=0)

As I mentioned in my earlier comment, if you use Quang Hoang's approach, you will get rid of the extra header rows, but all columns will be imported as object instead of integers. This could be more work if you have lots of columns of different datatypes. One solution might then be to export it back to csv and then import it back again...