从另一个文件中的一个文件中的数学上减去值

Question

我有两个文件，每个文件相同的大小（100x12），包含数字值，均由逗号分隔。

文件1的示例输出：

-14.99,-15.6,8.0 ->
-9.0,34.87,98.98 ->
(and so on)

文件2的示例输出：

-15.99,-18.6,8.00 ->
-3.0,34.34,-98.88 ->
(and so on)

我尝试过：

awk '{getline t<"file1"; print $0-t}' file2

但是，这仅减去第一列。如何将其扩展到数值从file2/column2？

中减去file1/column1。

我将开放使用熊猫执行此操作。预先感谢您！

Answer 1

首先数据

file1 = """
-14.99,-15.6,8.0 ->
-9.0,34.87,98.98 ->
"""
file2 = """
-15.99,-18.6,8.00 ->
-3.0,34.34,-98.88 ->
"""
from io import StringIO # faking file on disk

熊猫答案。

import pandas as pd
converter = {2: lambda s: float(s.split(' ')[0])}
df1 = pd.read_csv(StringIO(file1), header=None, converters=converter)
df2 = pd.read_csv(StringIO(file2), header=None, converters=converter)
(df1-df2).to_csv('pddiff12.csv', header=False, index=False)    

或用纯Python滚动。

# cmt 1 -> indent under with-statement
def read_csv(file_name):
    #with open('file_name', 'rt') as f1: # uncomment when reading from disk
    f1 = StringIO(file_name) # comment out when reading from disk
    rows = [r for r in f1.readlines() if r.strip()] # cmt 1
    crunch = lambda row: [float(r) for r in row.split(',')]
    rows = [crunch(r.split(' ')[0]) for r in rows]
    return rows
data1 = read_csv(file1)
data2 = read_csv(file2)
diff = []
for row1, row2 in zip(data1, data2):
    diff.append([i-j for i, j in zip(row1, row2)])
with open('diff12.csv', 'wt') as d12:
    for row in diff:
        d12.write(', '.join((str(v) for v in row)) + 'n')

pandas肯定最容易阅读和工作，尽管如果倾向于避免这种情况，这是一个明显的依赖性。在这种情况下，我认为我不会。

Answer 2

我的头顶 - 所以检查语法！

import numpy as np
with open("file1.txt, "r") as f1:
    with open("file2.txt, "r") as f2:
       array1 = np.asarray(f1.read().split(','))
       array2 = np.asarray(f2.read().split(','))
       result = array1 - array2
       print([x for x in result])

Answer 3

您可以尝试使用Unix实用程序。

paste file1.txt file2.txt | 
     awk -F"[,t]"  -v OFS="," ' { for(i=1;i<4;i++) { $i=$(i+3)-$i } print $1,$2,$3 } '

用给定的输入

$ cat halletwx1.txt
-14.99,-15.6,8.0
-9.0,34.87,98.98
$ cat halletwx2.txt
-15.99,-18.6,8.00
-3.0,34.34,-98.88
$ paste halletwx1.txt halletwx2.txt | awk -F"[,t]"  -v OFS="," ' { for(i=1;i<4;i++) { $i=$(i+3)-$i } print $1,$2,$3 } '
-1,-3,0
6,-0.53,-197.86
$

Answer 4

awk：

$ awk '
NR==FNR {                # hash file1 values to a
    for(i=1;i<=NF;i++)
        a[FNR][i]=$i
    next
}{                       # process file2, subtract values from file1 respectives
    for(i=1;i<=NF;i++)
        $i=$i-a[FNR][i]
}1' file1 file1

输出：

-1,-3,8.00 ->
6,-0.53,-98.88 ->