loadtxt()函数的糟心历程

原计划：导入一个csv文件，然后算出平均值

import numpy as np

c=np.loadtxt('d:\python36\data.csv', delimiter=',', usecols=(2,), unpack=True)

print ("median =", np.median(c))

---------------------------

报错：OSerror: d:\python36\data.csv not found

找了半天原因：原来是data.csv应该是data_1.csv

后来以为编码问题

import numpy as np

import codecs

myfile='d:\python36\data_1.csv'

filecp = codecs.open(myfile, encoding='utf-8')

c=np.loadtxt(filecp, skiprows=2,dtype=int)

print ("median =", np.median(c))

---------------------------

报错：ValueError:  could not convert string to float:’12,1,2,12’

搞了半天，没找到原因

去掉dtype=int后，报错IndexError:list index out of range

最终：

import numpy as np

import codecs

myfile='d:\python36\data_1.csv'

filecp = codecs.open(myfile, encoding='utf-8')

c=np.loadtxt(filecp, skiprows=2,delimiter=’,’, usecols=(2,),dtype=float)

print ("median =", np.median(c))

搞定，IndexError问题出在delimiter=’,’；而ValueError问题出在dtype=int和delimiter=’,’，因为没有delimiter=’,’隔开，就没法转换数据类型