Python numpy genfromtxt列

我在numpy genfromtxt上遇到了一个看似愚蠢而简单的问题。这是我的代码的（非常）简化版本：

import numpy as np

in_file_1 = raw_input ('enter name of template file to be scaled:\n')
spec_1 = np.genfromtxt(in_file_1, delimiter = [8,24], dtype =float)
print spec_1

我正在阅读的文件非常简单，两列没有标题等：

6392.01   0.90286163
6392.05   0.88731778
6392.09   0.87789831
6392.13   0.87716535
6392.16   0.88523003
6392.20   0.90948176
6392.24   0.93056874
6392.28   0.95782283
6392.32   0.98056805
6392.36   0.99623797
6392.39   0.99458828
6392.43   0.9848269
6392.47   0.96011146
6392.51   0.92864767

当我在python命令行上使用genfromtxt阅读上述内容时，它给出了我所期望的两列数组：

>>> import numpy as np
>>> in_file_1 = raw_input ('enter name of template file to be scaled:\n')
enter name of template file to be scaledl_1714650_052_no_head.txt
>>> spec_1 = np.genfromtxt(in_file_1, delimiter = [8,24], dtype =float)
>>> spec_1
array([[  6.39201000e+03,   9.02861630e-01],
       [  6.39205000e+03,   8.87317780e-01],
       [  6.39209000e+03,   8.77898310e-01],
       [  6.39213000e+03,   8.77165350e-01],
       [  6.39216000e+03,   8.85230030e-01],
       [  6.39220000e+03,   9.09481760e-01],
       [  6.39224000e+03,   9.30568740e-01],
       [  6.39228000e+03,   9.57822830e-01],
       [  6.39232000e+03,   9.80568050e-01],
       [  6.39236000e+03,   9.96237970e-01],
       [  6.39239000e+03,   9.94588280e-01],
       [  6.39243000e+03,   9.84826900e-01],
       [  6.39247000e+03,   9.60111460e-01],
       [  6.39251000e+03,   9.28647670e-01]])
>>> 

但当我在本文顶部以脚本（read_test.py）的形式运行它时，它会返回一列字符串：

[scrooge:Acc_cont_removal/All_stars/Test] darryl% python read_test.py 
enter name of template file to be scaled:
l_1714650_052_no_head.txt
[[  6.39201000e+03   9.02861630e-01]
 [  6.39205000e+03   8.87317780e-01]
 [  6.39209000e+03   8.77898310e-01]
 [  6.39213000e+03   8.77165350e-01]
 [  6.39216000e+03   8.85230030e-01]
 [  6.39220000e+03   9.09481760e-01]
 [  6.39224000e+03   9.30568740e-01]
 [  6.39228000e+03   9.57822830e-01]
 [  6.39232000e+03   9.80568050e-01]
 [  6.39236000e+03   9.96237970e-01]
 [  6.39239000e+03   9.94588280e-01]
 [  6.39243000e+03   9.84826900e-01]
 [  6.39247000e+03   9.60111460e-01]
 [  6.39251000e+03   9.28647670e-01]]
[scrooge:Acc_cont_removal/All_stars/Test] darryl% 

我已经尝试了各种分隔符的配置，但是我不知道发生了什么，几天前这段代码运行良好。使用PythonV2.7在OSX（Lion）上运行

任何想法都受到了感激。

---

Darryl

你得到的正是你应该得到的答案

>>> numpy.genfromtxt('testfile', delimiter=[8,24], dtype = float)
array([[  6.39201000e+03,   9.02861630e-01],
       [  6.39205000e+03,   8.87317780e-01],
       [  6.39209000e+03,   8.77898310e-01],
       [  6.39213000e+03,   8.77165350e-01],
       [  6.39216000e+03,   8.85230030e-01],
       [  6.39220000e+03,   9.09481760e-01],
       [  6.39224000e+03,   9.30568740e-01],
       [  6.39228000e+03,   9.57822830e-01],
       [  6.39232000e+03,   9.80568050e-01],
       [  6.39236000e+03,   9.96237970e-01],
       [  6.39239000e+03,   9.94588280e-01],
       [  6.39243000e+03,   9.84826900e-01],
       [  6.39247000e+03,   9.60111460e-01],
       [  6.39251000e+03,   9.28647670e-01]])
>>> print _
[[  6.39201000e+03   9.02861630e-01]
 [  6.39205000e+03   8.87317780e-01]
 [  6.39209000e+03   8.77898310e-01]
 [  6.39213000e+03   8.77165350e-01]
 [  6.39216000e+03   8.85230030e-01]
 [  6.39220000e+03   9.09481760e-01]
 [  6.39224000e+03   9.30568740e-01]
 [  6.39228000e+03   9.57822830e-01]
 [  6.39232000e+03   9.80568050e-01]
 [  6.39236000e+03   9.96237970e-01]
 [  6.39239000e+03   9.94588280e-01]
 [  6.39243000e+03   9.84826900e-01]
 [  6.39247000e+03   9.60111460e-01]
 [  6.39251000e+03   9.28647670e-01]]

这只是打印和获得结果表示之间的区别

>>> result = numpy.genfromtxt('testfile', delimiter=[8,24], dtype = float)
>>> repr(result)
'array([[  6.39201000e+03,   9.02861630e-01],\n       [  6.39205000e+03,   8.87317780e-01],\n       [  6.39209000e+03,   8.77898310e-01],\n       [  6.39213000e+03,   8.77165350e-01],\n       [  6.39216000e+03,
  8.85230030e-01],\n       [  6.39220000e+03,   9.09481760e-01],\n       [  6.39224000e+03,   9.30568740e-01],\n       [  6.39228000e+03,   9.57822830e-01],\n       [  6.39232000e+03,   9.80568050e-01],\n
[  6.39236000e+03,   9.96237970e-01],\n       [  6.39239000e+03,   9.94588280e-01],\n       [  6.39243000e+03,   9.84826900e-01],\n       [  6.39247000e+03,   9.60111460e-01],\n       [  6.39251000e+03,   9.2864
7670e-01]])'

>>> str(result)
'[[  6.39201000e+03   9.02861630e-01]\n [  6.39205000e+03   8.87317780e-01]\n [  6.39209000e+03   8.77898310e-01]\n [  6.39213000e+03   8.77165350e-01]\n [  6.39216000e+03   8.85230030e-01]\n [  6.39220000e+03
 9.09481760e-01]\n [  6.39224000e+03   9.30568740e-01]\n [  6.39228000e+03   9.57822830e-01]\n [  6.39232000e+03   9.80568050e-01]\n [  6.39236000e+03   9.96237970e-01]\n [  6.39239000e+03   9.94588280e-01]\n [
 6.39243000e+03   9.84826900e-01]\n [  6.39247000e+03   9.60111460e-01]\n [  6.39251000e+03   9.28647670e-01]]'

---

请注意，str版本没有逗号，而repr有逗号？

看起来不像一列字符串。事实上，它看起来几乎与正确的输出相同。但是当我引用它时，即：a=spec_1[0:1]它不返回值，我可以引用行（即a=spec[1]），它返回整行，但不返回单个元素。

spec_1[1]

应该返回整行，如果需要单个元素，可以使用

spec_1[1,0]

，

spec_1[2,1]

等。此外，如果在

test_read.py

的末尾添加

print spec_1.dtype、spec_1.shape

，您会得到什么输出？我得到

float64（14，2）

。