在 Pandas 中，如何使用变量名来表示行索引以获取可用作标题行的字符串？

Question

I'm trying to clean an excel file that has some random formatting. The file has blank rows at the top, with the actual column headings at row 8. I've gotten rid of the blank rows, and now want to use the row 8 string as the true column headings in the dataframe.

I use this code to get the position of the column headings by searching for the string 'Destination' in the whole dataframe, and then take the location of the True value in the Boolean mask to get the list for renaming the column headers:

boolmsk=df.apply(lambda row: row.astype(str).str.contains('Destination').any(), axis=1)
print(boolmsk)
hdrindex=boolmsk.index[boolmsk == True].tolist()
print(hdrindex)
hdrstr=df.loc[7]
print(hdrstr)
df2=df.rename(columns=hdrstr)

However when I try to use hdrindex as a variable, I get errors when the second dataframe is created (ie when I try to use hdrstr to replace column headings.)

boolmsk=df.apply(lambda row: row.astype(str).str.contains('Destination').any(), axis=1)
print(boolmsk)
hdrindex=boolmsk.index[boolmsk == True].tolist()
print(hdrindex)
hdrstr=df.loc[hdrindex]
print(hdrstr)
df2=df.rename(columns=hdrstr)

How do I use a variable to specify an index, so that the resulting list can be used as column headings?

Answer 1

I assume your indicator of actual header rows in dataframe is string "destination". Lets find where it is:

start_tag = df.eq("destination").any(1)

We'll keep the number of the index of first occurrence of word "destination" for further use:

start_row = df.loc[start_tag].index.min()

Using index number we will get list of values in the "header" row:

new_col_names = df.iloc[start_row].values.tolist()

And here we can assign new column names to dataframe:

df.columns = new_col_names

From here you can play with new dataframe, actual column names and proper indexing:

df2 = df.iloc[start_row+1:].reset_index(drop=True)