使用Pandas从文件中读取分层的ASCII表 [英] Reading from file a hierarchical ascii table using Pandas

问题描述

我在这里的第一篇文章，希望不会太长或太详细.

My very-first post here, I hope it's not too long or detailed.

当我尝试读取和解释以下ascii表(从一个更大的表中提取的简单内容)时，我遇到了一个问题.

I have an issue when I try to read and interpret the following ascii table (simple extract from a much bigger one).

假设文件名为"test.txt":

Let's say the file is called "test.txt":

      A         B          C             D             E    
0       992    CEN/4 -2.657293E+00 -3.309567E+01  4.697218E-01
                1291 -3.368449E+00  7.837483E+00  2.311393E+00
                 800 -3.530800E+00 -7.392188E+01 -1.401380E+00
                 801 -1.952177E+00 -7.392114E+01 -1.367195E+00
                1290 -1.777301E+00  7.838229E+00  2.345850E+00
0       994    CEN/4  7.270955E+00 -6.637891E+01 -1.293553E+01
                1110  5.816999E+00 -3.981042E+01 -1.504738E+01
                 785  5.535329E+00 -9.246339E+01 -1.061554E+01
                 786  8.625161E+00 -9.163719E+01 -1.092563E+01
                1109  9.080517E+00 -4.059749E+01 -1.523589E+01

使用python 2.7和pandas(0.9.1)，我可以阅读以下内容:

Using python 2.7 and pandas (0.9.1), I can read it as follows:

>>> r=pd.read_fwf('test.txt', widths=(3, 8, 9, 14, 14, 14),skiprows=1, header=None)
>>> print r

   X0   X1     X2        X3         X4         X5
0   0  992  CEN/4 -2.657293 -33.095670   0.469722
1 NaN  NaN   1291 -3.368449   7.837483   2.311393
2 NaN  NaN    800 -3.530800 -73.921880  -1.401380
3 NaN  NaN    801 -1.952177 -73.921140  -1.367195
4 NaN  NaN   1290 -1.777301   7.838229   2.345850
5   0  994  CEN/4  7.270955 -66.378910 -12.935530
6 NaN  NaN   1110  5.816999 -39.810420 -15.047380
7 NaN  NaN    785  5.535329 -92.463390 -10.615540
8 NaN  NaN    786  8.625161 -91.637190 -10.925630
9 NaN  NaN   1109  9.080517 -40.597490 -15.235890

尝试将其直接读取为层次表":

Trying to read it directly as a "hierarchical table":

>>> r=pd.read_fwf('test.txt', widths=(3, 8, 9, 14, 14, 14), skiprows=1, index_col=[0,1,2], header=None)
>>> print r

                    X3         X4         X5
X0 X1  X2                                   
0  992 CEN/4 -2.657293 -33.095670   0.469722
   994 1291  -3.368449   7.837483   2.311393
       800   -3.530800 -73.921880  -1.401380
       801   -1.952177 -73.921140  -1.367195
       1290  -1.777301   7.838229   2.345850
       CEN/4  7.270955 -66.378910 -12.935530
       1110   5.816999 -39.810420 -15.047380
       785    5.535329 -92.463390 -10.615540
       786    8.625161 -91.637190 -10.925630
       1109   9.080517 -40.597490 -15.235890

我的目标是获得:

>>> print r
                    X3         X4         X5
X0 X1  X2                                   
0  992 CEN/4 -2.657293 -33.095670   0.469722
       1291  -3.368449   7.837483   2.311393
       800   -3.530800 -73.921880  -1.401380
       801   -1.952177 -73.921140  -1.367195
       1290  -1.777301   7.838229   2.345850
   994 CEN/4  7.270955 -66.378910 -12.935530
       1110   5.816999 -39.810420 -15.047380
       785    5.535329 -92.463390 -10.615540
       786    8.625161 -91.637190 -10.925630
       1109   9.080517 -40.597490 -15.235890

是否有使用熊猫的简单方法，还是我必须在解析之前先按摩" ascii表以获得所需的信息? 预先感谢

Is there a straightforward way using pandas, or am I obliged to "massage" the ascii table before parsing to get what I want? Thanks in advance

推荐答案

看起来您需要在索引之前填充NA值.试试这个:

Looks like you need to fill the NA values before indexing. Try this:

r=pd.read_fwf('test.txt', widths=(3, 8, 9, 14, 14, 14),skiprows=1, header=None)
r=r.fillna(method='pad')
r=r.set_index(['X0','X1','X2'])

这篇关于使用Pandas从文件中读取分层的ASCII表的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！