insert value string into list where list in each cell of column-CodePudding

My input:

df=(pd.DataFrame({'items':['foo','bar','xyz'],
                  'path':['c/folderOne/folderTwo/folderThree','c/folder1/folder2/folder3','c/folderO/folderT/folderTh'],
}))

I trying insert in column path some "fix" after 'c/... I want add new substring such as NEWPATH into list and re-join again, but I ran unexpected problem.
That my code:

df['path'] = df['path'].map(lambda x: x.split("/"))
df['path'] = df['path'].map(lambda x: x.insert(1,'NEWPATH'))
df['path'] = df['path'].map(lambda x: x.join('/'))

Code return None in column path, when I expected something like this, after second line of my code:
in column df['path'] : ['c','NEWPATH','folderOne'....] etc. for each row(cell) of my column.

CodePudding user response：

Looks like you should better use a regex here:

NEWPATH = 'abc'
df['path2'] = df['path'].str.replace('^([^/] /)', rf'\1{NEWPATH}/')

NB. assigning to a new column here for clarity

output:

  items                               path                                  path2
0   foo  c/folderOne/folderTwo/folderThree  c/abc/folderOne/folderTwo/folderThree
1   bar          c/folder1/folder2/folder3          c/abc/folder1/folder2/folder3
2   xyz         c/folderO/folderT/folderTh         c/abc/folderO/folderT/folderTh

inserting after the nth directory:

NEWPATH = 'abc'
df['path2'] = df['path'].str.replace('^(^(?:[^/] /){3})', rf'\1{NEWPATH}/')

NEWPATH = 'abc'
N = 3
df['path2'] = df['path'].str.replace(f'^(^(?:[^/] /){{{N}}})', rf'\1{NEWPATH}/')

output:

  items                               path                                  path2
0   foo  c/folderOne/folderTwo/folderThree  c/folderOne/folderTwo/abc/folderThree
1   bar          c/folder1/folder2/folder3          c/folder1/folder2/abc/folder3
2   xyz         c/folderO/folderT/folderTh         c/folderO/folderT/abc/folderTh

CodePudding user response：

Problem is insert working inplace, so None is returned. You can add list to splitted x and then join:

def f(x):
    s = x.split("/")
    return '/'.join(s[:2]   ['NEWPATH']   s[2:])
    
df['path'] = df['path'].map(f)

print (df)
  items                                       path
0   foo  c/folderOne/NEWPATH/folderTwo/folderThree
1   bar          c/folder1/NEWPATH/folder2/folder3
2   xyz         c/folderO/NEWPATH/folderT/folderTh

CodePudding user response：

You can use a map as well - the index 1 in the lambda here indicates the position where you want to insert

df['path'] = df['path'].str.split('/').map(lambda x: x[:1]   ['NEWPATH']   x[1:]).str.join('/')