使用Pip和Easy_install进行Pandas安装

时间:2017-04-17 15:42:56

标签: python pandas numpy

我的熊猫有问题。首先,我尝试在cd C:\ Python27 \ Scrips ......上轻松安装。 easy_install.exe numpyeasy_install.exe pandas,但只有numpy工作,大熊猫没有,所以我用pip安装。它对两者都有效。但回到我的代码,当我测试它时,

继承我的代码:

import fileinput
import csv
import numpy
import pandas
#records = pe.iget_records(file_name="test.xlxs")
records = pd.read_csv('test.csv', header=None, nrows=5)

#To Write ni





cho = raw_input("\nStart Forecaster on file?:<1/0>")

if cho == 1:

  for record in records:
     rem = df.iloc([i], [0])
     print(rem)
     sold1 = df.iloc([i], [1])
     print(sold1)
     sold2 = df.iloc([i], [2])
     print(sold2)

     rem = int(rem)
     sold1 = int(sold1)
     sold2 = int(sold2)

     result = forecast(rem,sold1,sold2)
     print(result)
     df.set_value([i], [4], result)
     pd.to_csv('test.csv')
  print "Forecast Complete! Please check the file!"


else:
  quit() 

但是我收到了这个错误:

   Traceback (most recent call last):
      File "tester.py", line 4, in <module>
        import pandas
      File "C:\Python27\lib\site-packages\pandas\__init__.py", line 25, in <module>
        from pandas import hashtable, tslib, lib
      File "pandas\src\numpy.pxd", line 157, in init pandas.hashtable (pandas\hashtable.c:40866)
    ValueError: numpy.dtype has the wrong size, try recompiling. Expected 52, got 56

1 个答案:

答案 0 :(得分:0)

您需要删除以前安装的pandasnumpy,因为您连续安装了太多次,导致库混淆。删除后,使用pandas重新安装numpypip,如下所示:

pip install numpy
pip install pandas