如何在 Python 中将 2 列 ASCII 文件数据正确拆分为多列数据？

Question

我有一个结构如下的文本文件：

0   1.23
1   2.76
2   2.46
3   6.23

0   1.33
1   2.57
2   2.87
3   5.34

.
.
.

我想安排一个新文件，例如：

0   1.23   1.33  ...
1   2.76   2.57  ...
2   2.46   2.87  ...
3   6.23   5.34  ...

我可以用一种非常原始的方式来做到这一点：

# Number of data group
numberofdatagroup = 5
# Number of data in each group
data = 4


arr = [[0 for col in range(2*numberofdatagroup)] for row in range(data)]
f = open(file, 'r')
lines = f.readlines()
f.close()
a=0
for i in range(0, numberofdatagroup, 1):
   b = 0
   for a in range (0, data, 1):
      fields = lines[a].split()
      arr[b][2*i] = fields[0]
      arr[b][2*i+1] = fields[1]
      b = b + 1
   a = a + 2

# writing to output file
f = open(output, 'w+')
stringline = ""

for i in range(0, data, 1):
  stringline = stringline + arr[i][0] + " " + arr[i][1] + " "
  for j in range(1, numberofdatagroup, 1):
     stringline = stringline + arr[i][2*j+1] + " "
  f.write(stringline + "\n")
  stringline = ""

f.close()

但是，它并不总是有效。对空行非常敏感。有没有办法用更聪明的方式来制作它？

Answer 1

这是一个如何将文件读入 Pandas DataFrame 的示例：

import pandas as pd

current, all_groups = [], []
with open('data.txt', 'r') as f_in:
    for line in map(str.strip, f_in):
        if line == "" and current:
            all_groups.append(pd.DataFrame(current)[1])
            current = []
        else:
            current.append(line.split(maxsplit=1))

if current:
    all_groups.append(pd.DataFrame(current)[1])

final_df = pd.concat(all_groups, axis=1)
final_df.columns = range(len(final_df.columns))

print(final_df)

印刷：

      0     1
0  1.23  1.33
1  2.76  2.57
2  2.46  2.87
3  6.23  5.34

---

编辑：没有pandas库：

current, all_groups = [], []
with open("data.txt", "r") as f_in:
    for line in map(str.strip, f_in):
        if line == "" and current:
            all_groups.append(current)
            current = []
        else:
            current.append(line.split(maxsplit=1))

if current:
    all_groups.append(current)

for g in zip(*all_groups):
    print('{} {} {}'.format(g[0][0], g[0][1], ' '.join(v for _, v in g[1:])))