Question

我有一个* .txt而以下配置：一个长标头和前面的数据。见下文

field1, field2, field3
field4, field5, field 6, field7, field8
field9, fiel10
field11, field12

1, 1.1, 1o.1
2, 0.5, 15
3, 0, 8.3
4, 2.1, 7.8
.. 
..

这是我制作的代码。为了从头部保存值，我创建了一个名为＆＃34; header＆＃34;的字典。

header={}
count=1
with open('file.txt') as f:
   while count<4:
      line = f.readline()
      if count==1:
         header['field1]=line.split(',')[0]
         header['field2]=line.split(',')[1]          
         header['field3]=line.split(',')[2]
      if count==2:
         header['field4]=line.split(',')[0]
         header['field5]=line.split(',')[1]          
         header['field6]=line.split(',')[2]
         header['field7]=line.split(',')[3]          
         header['field8]=line.split(',')[4]
      if count==3:
         header['field9]=line.split(',')[0]
         header['field10]=line.split(',')[1]          
      if count==4:
         header['field11]=line.split(',')[0]
         header['field12]=line.split(',')[1]          

#Read the full data into dataframe
df=  pd.read_csv('file.txt',skiprows=4,names=['Col1','Col2','Col3])

然而，我认为这样做并不是很有效也不优雅。我将不胜感激使用I / O文件指针或Pandas的简单版本。感谢

Answer 1

迭代标题行，拆分它们，然后遍历行enries：

header = {}
with open('file.txt') as fobj:
    counter = 1
    for line in fobj:
        # assuming empty line between multi-line header and data
        if not line.strip():
            break
        for entry in line.split(','):
            header['field{}'.format(counter)] = entry.strip()
            counter += 1

import pprint    
pprint.pprint(header)

输出：

{'field1': 'field1',
 'field10': 'fiel10',
 'field11': 'field11',
 'field12': 'field12',
 'field2': 'field2',
 'field3': 'field3',
 'field4': 'field4',
 'field5': 'field5',
 'field6': 'field 6',
 'field7': 'field7',
 'field8': 'field8',
 'field9': 'field9'}

将多字头读入字典，将数据读入DataFrame

1 个答案: