转置将另一行/列添加到 dataframe

Question

When trying to transpose the following table:

Sample	input_test	input_test2	input_test3	ip_test	ip_test2	ip_test3
tRNA	5	5	5	1	1	1
CDS	9330	9330	9330	26680	26680	26680
3utr	2525	2525	2525	3810	3810	3810
5utr	1966	1966	1966	5006	5006	5006
5ss	960	960	960	1972	1972	1972
3ss	932	932	932	1989	1989	1989
proxintron	1221	1221	1221	1228	1228	1228
distintron	6450	6450	6450	4744	4744	4744
noncoding_exon	1477	1477	1477	1302	1302	1302
noncoding_5ss	25	25	25	35	35	35
noncoding_3ss	16	16	16	29	29	29
noncoding_proxintron	68	68	68	80	80	80
noncoding_distintron	629	629	629	441	441	441
allexonic	15298	15298	15298	36798	36798	36798
allintronic	10301	10301	10301	10518	10518	10518
all	27131	27131	27131	49060	49060	49060

I get this:

0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
Sample	tRNA	CDS	3utr	5utr	5ss	3ss	proxintron	distintron	noncoding_exon	noncoding_5ss	noncoding_3ss	noncoding_proxintron	noncoding_distintron	allexonic	allintronic	all
input_test	5	9330	2525	1966	960	932	1221	6450	1477	25	16	68	629	15298	10301	27131
input_test2	5	9330	2525	1966	960	932	1221	6450	1477	25	16	68	629	15298	10301	27131
input_test3	5	9330	2525	1966	960	932	1221	6450	1477	25	16	68	629	15298	10301	27131
ip_test	1	26680	3810	5006	1972	1989	1228	4744	1302	35	29	80	441	36798	10518	49060
ip_test2	1	26680	3810	5006	1972	1989	1228	4744	1302	35	29	80	441	36798	10518	49060
ip_test3	1	26680	3810	5006	1972	1989	1228	4744	1302	35	29	80	441	36798	10518	49060

Why is there an additional row with numbers on top? This is driving me nuts, I've tried so many options with index=False , df_merged.set_index('Sample',inplace=True) , etc.

This is the code:

# transpose

    df_merged = pd.read_csv(str(save_path) + "all_stats_matrix.csv")

    df_merged = df_merged.T

    df_merged.to_csv(str(save_path) + "all_stats_matrix.csv")

Answer 1

You could alternatively if your goal is to have the first row as headers instead of the numbers. Use this:

df_merged = pd.read_csv(str(save_path) + "all_stats_matrix.csv")

df_merged = df_merged.T

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

df_merged.to_csv(str(save_path) + "all_stats_matrix.csv")

It is not the most clean way, but should get the job done.

Answer 2

How about: df_merged.set_index('Sample').transpose() ?