PYTHON：删除一些行后将一列拆分为多列

Question

我是 Python 的新手，我正在尝试清理一些数据。我已将 link 附加到数据文件（两个选项卡：原始数据和预期结果）。请帮忙！

我想做什么：

删除第 1-23 行
使用“-”作为分隔符将列 B 拆分为多列
为新列指定列名
保留数字列

Link 到原始数据（第一个选项卡）和预期结果（第二个选项卡）： https://www.dropbox.com/s/kjgtwoelq21eetw/Example2.xlsx?dl=0

我目前拥有的：

import numpy as np
data_xls=pd.read_excel("Example2.xlsx", index_col=None).fillna('')
data_xls = data_xls.iloc[22:]
data_xls.rename(columns=data_xls.iloc[0]).drop(data_xls.index[0])
data_xls['Internal Link Tracking (non-promotions) - ENT (c20)'].str.split('-', expand=True)

writer = pd.ExcelWriter('Output2.xlsx')
data_xls.to_excel(writer, 'O1', index=False)
writer.save()

非常感谢您的帮助！泰

Answer 1

使用：

# Read the excel file with sheet_name='Raw data' and skiprows=23 which are not necessary
data_xls = pd.read_excel("Example2.xlsx", sheet_name='Raw data', skiprows=23)

# Create the dummy columns names which are similar to desired output column
dummy_col_names = ['Internal Link Tracking (non','Campaign Name','Creative','Action','Action 2']
# Use str.split with expand=True to create a dataframe
dummy_df = data_xls['Internal Link Tracking (non-promotions) - ENT (c20)'].str.split('-',expand = True)
# Rename columns as per dummy column list
dummy_df.columns = dummy_col_names

# Drop the column which is not necessary
data_xls.drop('Internal Link Tracking (non-promotions) - ENT (c20)', axis=1, inplace=True)

# Use pd.concat along axis=1 to concat both data_xls and dummy_df along columns
data_xls = pd.concat((data_xls,dummy_df),sort=False,axis=1)

# To preserve oreder similar to desired output column use the following code
col_names = data_xls.columns.tolist()
data_xls = data_xls[col_names[:1]+dummy_col_names+col_names[1:-5]]

Answer 2

使用 pandas

将一列拆分为 2 列

d = pd.read_csv('file.csv')

   col_1
    "val1-val2"
    "valA-valB"

df = pd.DataFrame(d.col_1.str.split("-",1).tolist(),columns = ['A','B'])

      A     B
0  val1  val2
1  valA  valB

Answer 3

试试这个：

1.)删除第1-23行

df = pd.read_excel('/home/mayankp/Downloads/Example2.xlsx', sheet_name=0, index_col=None, header=None, skiprows=23)

2.) 使用“-”作为分隔符将列 B 拆分为多列 和 3.) 将列名称分配给新列

这两个步骤可以一次完成：

sub_df = df[1].str.split('-', expand=True).rename(columns = lambda x: "string"+str(x+1))

In [179]: sub_df
Out[179]: 
                       string1       string2             string3      string4     string5
1                           us      campaign            article1   scrolldown  findoutnow
2                           us      campaign            article1  scrollright        None
3                           us      campaign            article1   findoutnow        None
4                           us      campaign  payablesmanagement   findoutnow        None

上面是样本在 - 上拆分后的样子。

现在从 df 中删除实际列并在其中插入这些新列：

df = df.drop(1, axis=1)
df = pd.concat([df,sub_df], axis=1)

4.)保留数字列

其余的列已经完好无损。无需更改。

如果有帮助请告诉我。

PYTHON：删除一些行后将一列拆分为多列

PYTHON: Splitting One Column into Multiple After Deleting some rows

python

split

delete-row

pandas