使用Pandas读取数据（.dat文件）1

Question

如何使用Pandas读取以下（两列）数据（来自.dat文件）

TIME                      XGSM
2004 006 01 00 01 37 600  1
2004 006 01 00 02 32 800  5
2004 006 01 00 03 28 000  8
2004 006 01 00 04 23 200  11
2004 006 01 00 05 18 400  17

列分隔符为（至少）2个空格。

我试过了

df = pd.read_table("test.dat", sep="\s+", usecols=['TIME', 'XGSM'])
print df

但有指纹

   TIME  XGSM
   2004     6
   2004     6
   2004     6
   2004     6
   2004     6

Answer 1

可以将参数usecols与列的顺序一起使用：

import pandas as pd
from pandas.compat import StringIO

temp=u"""TIME             XGSM
2004 006 01 00 01 37 600  1
2004 006 01 00 02 32 800  5
2004 006 01 00 03 28 000  8
2004 006 01 00 04 23 200  11
2004 006 01 00 05 18 400  17"""
#after testing replace StringIO(temp) to filename
df = pd.read_csv(StringIO(temp), 
                 sep="\s+", 
                 skiprows=1, 
                 usecols=[0,7], 
                 names=['TIME','XGSM'])

print (df)
   TIME  XGSM
0  2004     1
1  2004     5
2  2004     8
3  2004    11
4  2004    17

编辑：

您可以使用分隔符regex-2和更多空格，然后添加engine='python'，因为警告：

ParserWarning: Falling back to the 'python' engine because the 'c' engine does not support regex separators (separators > 1 char and different from '\s+' are interpreted as regex); you can avoid this warning by specifying engine='python'.

import pandas as pd
from pandas.compat import StringIO

temp=u"""TIME              XGSM
2004 006 01 00 01 37 600   1
2004 006 01 00 02 32 800   5
2004 006 01 00 03 28 000   8
2004 006 01 00 04 23 200   11
2004 006 01 00 05 18 400   17"""
#after testing replace StringIO(temp) to filename
df = pd.read_csv(StringIO(temp), sep=r'\s{2,}', engine='python')

print (df)
                       TIME  XGSM
0  2004 006 01 00 01 37 600     1
1  2004 006 01 00 02 32 800     5
2  2004 006 01 00 03 28 000     8
3  2004 006 01 00 04 23 200    11
4  2004 006 01 00 05 18 400    17

Answer 2

也可以尝试pd.read_fwf()（将固定宽度格式化行的表读取到DataFrame）：

import pandas as pd
from io import StringIO

pd.read_fwf(StringIO("""TIME                      XGSM
2004 006 01 00 01 37 600  1
2004 006 01 00 02 32 800  5
2004 006 01 00 03 28 000  8
2004 006 01 00 04 23 200  11
2004 006 01 00 05 18 400  17"""), usecols = ["TIME", "XGSM"])

#   TIME    XGSM
#0  2004    1
#1  2004    5
#2  2004    8
#3  2004    11
#4  2004    17