Replacing Header with Top Row

Question

I currently have a dataframe that looks like this:

       Unnamed: 1    Unnamed: 2   Unnamed: 3  Unnamed: 4
0   Sample Number  Group Number  Sample Name  Group Name
1             1.0           1.0          s_1         g_1
2             2.0           1.0          s_2         g_1
3             3.0           1.0          s_3         g_1
4             4.0           2.0          s_4         g_2

I'm looking for a way to delete the header row and make the first row the new header row, so the new dataframe would look like this:

    Sample Number  Group Number  Sample Name  Group Name
0             1.0           1.0          s_1         g_1
1             2.0           1.0          s_2         g_1
2             3.0           1.0          s_3         g_1
3             4.0           2.0          s_4         g_2

I've tried stuff along the lines of if 'Unnamed' in df.columns: then make the dataframe without the header

df.to_csv(newformat, header=False, index=False)

but I don't seem to be getting anywhere.

Answer 1

new_header = df.iloc[0] #grab the first row for the header
df = df[1:] #take the data less the header row
df.columns = new_header #set the header row as the df header

Answer 2

The dataframe can be changed by just doing

df.columns = df.iloc[0]
df = df[1:]

Then

df.to_csv(path, index=False) 

Should do the trick.

Answer 3

If you want a one-liner, you can do:

df.rename(columns=df.iloc[0]).drop(df.index[0])

Answer 4

Another one-liner using Python swapping:

df, df.columns = df[1:] , df.iloc[0]

This won't reset the index

Although, the opposite won't work as expected df.columns, df = df.iloc[0], df[1:]

Answer 5

@ostrokach answer is best. Most likely you would want to keep that throughout any references to the dataframe, thus would benefit from inplace = True.
df.rename(columns=df.iloc[0], inplace = True) df.drop([0], inplace = True)

Answer 6

Here's a simple trick that defines column indices "in place". Because set_index sets row indices in place, we can do the same thing for columns by transposing the data frame, setting the index, and transposing it back:

df = df.T.set_index(0).T

Note you may have to change the 0 in set_index(0) if your rows have a different index already.

Answer 7

Alternatively, we can do this when reading a file with pandas.

This case we can use,

pd.read_csv('file_path',skiprows=1)

When reading the file this will skip the first row and will set the column as the second row of the file.

Answer 8

For some reason, I had to do it this way:

df.columns = [*df.iloc[0]]
df = table[1:]

The part where I'm splitting the list into a list looks redundant, but otherwise, the headers still turn up as part of the actual table.

Answer 9

--another way to do this

df.columns = df.iloc[0]
df = df.reindex(df.index.drop(0)).reset_index(drop=True)
df.columns.name = None

    Sample Number  Group Number  Sample Name  Group Name
0             1.0           1.0          s_1         g_1
1             2.0           1.0          s_2         g_1
2             3.0           1.0          s_3         g_1
3             4.0           2.0          s_4         g_2

If you like it hit up arrow. Thanks

Answer 10

header = table_df.iloc[0]
table_df.drop([0], axis =0, inplace=True)
table_df.reset_index(drop=True)
table_df.columns = header
table_df

Answer 11

This seems like a task that may be needed more than once. I've taken rgalbo's answer and written a simple function that can be lifted and placed into any project.

def promote_df_headers(df):
    '''
    Takes a df and uses the first row as the header

    Parameters
    ----------
    df : DataFrame
        Any df with one or more columns.

    Returns
    -------
    df : DataFrame
        Input df with the first row removed and used as the column names.

    '''

    new_header = df.iloc[0] 
    df = df[1:] 
    df.columns = new_header
    df = df.reset_index(drop=True)

    return df

Answer 12

If you are starting with a list of lists

pd.DataFrame(input[1:], columns=input[0])

Answer 13

The best practice and Best OneLiner:

df.to_csv(newformat,header=1)

Notice the header value:

Header refer to the Row number(s) to use as the column names. Make no mistake, the row number is not the df but from the excel file(0 is the first row, 1 is the second and so on).

This way, you will get the column name you want and won't have to write additional codes or create new df.

Good thing is, it drops the replaced row.