合并单元格，在同一列，在同一个 df- Python

Question

我正在尝试将两个单元格合并在一起。这样做的原因是 'Chassis' 下的每个单位都应该是字母数字 (ABCD123456)，但是提供的 PO 偶尔会将最后一个数字移到下一行（该行上没有其他数据），使数据看起来像这样 Example I initially tried to create a statement that looked at the cell, confirmed it was less one number, then would look at the next cell, and merge the two. Never got that to even come close to manifesting any results. I then decided to replicate the data frame, shift the second data frame(so the missing number is on the same row), and merge them together. This is where I am now. Error Msg 这是我在 Python 中的第一个真正的代码，所以我相当确定我在做低效的事情所以一定要让我知道我可以改进的地方。

目前我有这个...

Col1	Chassis	Other Columns...	Other Columns 2...
Nan	ABCD12345	ABC	123
Nan	6	Nan	Nan
Nan	WXYZ987654	GHI	456
Nan	QRSTU654987	Nan	789
Nan	MNOP999999	XYZ	Nan

最终目标是这个...

Col1	Chassis	Other Columns...	Other Columns 2...
Nan	ABCD123456	ABC	123
Nan	WXYZ987654	GHI	456
Nan	QRSTU654987	Nan	789
Nan	MNOP999999	XYZ	Nan

import PyPDF2 as pdf2
import tabula as tb
import pandas as pd
import re
import csv
import os

os.listdir()
pd.set_option('display.max_columns', None)

#bring in pdf, remove first page, convert to csv
PO = 'PO.pdf'
pages = open(PO, 'rb')
readPDF = pdf2.PdfFileReader(pages)
totalpages = readPDF.numPages
x = '2-' + str(totalpages)
POCSV = tb.convert_into(PO, 'POCSV.csv', output_format = 'csv', pages = x)

#Convert column to string, create second data frame, shift said data frame up 1
df = pd.read_csv('POCSV.csv')
df['Chassis'] = df['Chassis'].astype(str)
dfshift = df.shift(-1)
dfshift.rename(columns=({'Chassis': 'Chassis Shifted'}), inplace = True,)
dfMerged = pd.concat([df, dfshift], axis=1)

#For each row combine rows, create new column
for ind, row in df.iterrows():
    dfMerged.loc[ind, 'Complete Chassis'] = row['Chassis'] + row["Chassis Shifted"]
    
print(dfMerged['Complete Chassis'])

Answer 1

为 Chassis 列创建一个虚拟组并合并该组的行：

# Convert 'NaN' string to pd.NA
df = df.replace('Nan', pd.NA)

cols = df.columns.difference(['Chassis'])
m = df[cols].any(1)
df = df.assign(Chassis=df.groupby(m.cumsum())['Chassis'] \
       .transform('sum')).loc[m].reset_index(drop=True)
print(df)

# Output
   Col1      Chassis Other Columns Other Columns 2
0  <NA>   ABCD123456           ABC             123
1  <NA>   WXYZ987654           GHI             456
2  <NA>  QRSTU654987          <NA>             789
3  <NA>   MNOP999999           XYZ            <NA>

设置：

import pandas as pd
import numpy as np

data = {'Col1': ['Nan', 'Nan', 'Nan', 'Nan', 'Nan'],
        'Chassis': ['ABCD12345', '6', 'WXYZ987654', 'QRSTU654987', 'MNOP999999'],
         'Other Columns': ['ABC', 'Nan', 'GHI', 'Nan', 'XYZ'],
         'Other Columns 2': ['123', 'Nan', '456', '789', 'Nan']}
df = pd.DataFrame(data)
print(df)

# Output
  Col1      Chassis Other Columns Other Columns 2
0  Nan    ABCD12345           ABC             123
1  Nan            6           Nan             Nan
2  Nan   WXYZ987654           GHI             456
3  Nan  QRSTU654987           Nan             789
4  Nan   MNOP999999           XYZ             Nan

合并单元格，在同一列，在同一个 df- Python

Merging cells, in the same column, in the same df- Python

python

merge

pandas

pypdf2

tabula-py