标签：
数据清洗，使用python数据清洗cvs里面带中文字符，意图是用字典对应中文字符，即key值是中文字符，value值是index，自增即可；利用字典数据结构没有重复key值的特性，把中文字符映射到了数值index。
python代码如下：（data数据时csv格式）
import csv
dict2 = {} #C
dict4 = {} #E
dict25 = {} #z
dict26 = {} #AA
dict27 = {} #AB
dict37 = {} #AL
dict38 = {} #AM
dict40 = {} #AO
dict41 = {} #AP
dict42 = {} #AQ
dict45 = {} #AT
dict49 = {} #AX
index = 0
flag = False
# print(row[2],dict[row[2]])
with open("E:/yuce/test.csv", ‘w+‘, newline=‘‘) as csv_file_write:
writer = csv.writer(csv_file_write)
with open(‘E:/yuce/b.csv‘, ‘r‘, newline=‘‘) as csv_file_read:
reader = csv.reader(csv_file_read)
for row in reader:
if(flag):
dict2[row[2]] = index
dict4[row[4]] = index
dict25[row[25]] = index
dict26[row[26]] = index
dict27[row[27]] = index
dict37[row[37]] = index
dict38[row[38]] = index
dict40[row[40]] = index
dict41[row[41]] = index
dict42[row[42]] = index
dict45[row[45]] = index
dict49[row[49]] = index
row[2] = dict2[row[2]]
row[4] = dict4[row[4]]
row[25] = dict25[row[25]]
row[26] = dict26[row[26]]
row[27] = dict27[row[27]]
row[37] = dict37[row[37]]
row[38] = dict38[row[38]]
row[40] = dict40[row[40]]
row[41] = dict41[row[41]]
row[42] = dict42[row[42]]
row[45] = dict45[row[45]]
row[49] = dict49[row[49]]
index = index + 1
writer.writerow(row)
flag = True
csv_file_read.close()
csv_file_write.close()
print(‘done!‘)
上例是真实的数据处理，有两百列属性，三万条数据的原始数据。其中包括中文字符，及缺失值，需要一步步清洗。
标签：
原文地址：http://www.cnblogs.com/rongyux/p/5404917.html
