如何将CSV文件的选定标题转换为Pandas数据框

Question

I have the following CSV data

id,gene,celltype,stem,stem,stem,bcell,bcell,tcell
id,gene,organs,bm,bm,fl,pt,pt,bm
id,gene,organs,stem1,stem2,stem3,b1,b2,t1
134,foo,about_foo,20,10,11,23,22,79
222,bar,about_bar,17,13,55,12,13,88

The first three lines are the header. What I want to do is to select line 1 and 3 and turn it into a data frame that looks like this:

Coln1 Coln2
stem  stem1
stem  stem2
stem  stem3
bcell b1
bcell b2
tcell t1

I am stuck with the following:

import pandas as pd
df = pd.read_csv("http://dpaste.com/00AWDBW.txt",header=None,index_col=[1,2]).iloc[:, 1:]

Answer 1

You can use parameters nrows and skiprows in read_csv :

import pandas as pd
import io

temp=u"""id,gene,celltype,stem,stem,stem,bcell,bcell,tcell
id,gene,organs,bm,bm,fl,pt,pt,bm
id,gene,organs,stem1,stem2,stem3,b1,b2,t1
134,foo,about_foo,20,10,11,23,22,79
222,bar,about_bar,17,13,55,12,13,88"""

#after testing replace io.StringIO(temp) to filename
df = pd.read_csv(io.StringIO(temp),header=None,index_col=[1,2], nrows=2, skiprows=[1])
df = df.ix[:, 1:].reset_index(drop=True).T
df.columns = ['Coln1', 'Coln2']
print df.reset_index(drop=True)

   Coln1  Coln2
0   stem  stem1
1   stem  stem2
2   stem  stem3
3  bcell     b1
4  bcell     b2
5  tell     t1

To select top 3 header into columns do this:

df = pd.read_csv(io.StringIO(temp),header=None,index_col=[1,2], nrows=3, skiprows=[4])
df = df.ix[:, 1:].reset_index(drop=True).T
df.columns = ['Coln1', 'Coln2','Coln3']
print df.reset_index(drop=True)