如何用某些字符替换列的开头和结尾 python dataframe

Question

I have a dataframe that looks like this:

 clients_x                 clients_y              coords_x               coords_y 
7110001002                7100019838    -23.63013,-46.704887  -23.657433,-46.744095   
7110001002                7100021875    -23.63013,-46.704887    -23.7729,-46.591366   
7110001002                0700245857    -23.63013,-46.704887      -23.7074,-46.5698 
[7110052941, 7110107795]  7100019838        -23.609,-46.6974  -23.657433,-46.744095
[7110052941, 7110107795]  7100021875        -23.609,-46.6974    -23.7729,-46.591366
[7110052941, 7110107795]  0700245857        -23.609,-46.6974       -23.7074,-46.569

What I want to do is for all values in clients_x column to start and end with "[ ]". Therefore, my expected output is this one:

 clients_x                 clients_y              coords_x               coords_y 
[7110001002]                7100019838    -23.63013,-46.704887  -23.657433,-46.744095   
[7110001002]                7100021875    -23.63013,-46.704887    -23.7729,-46.591366   
[7110001002]                0700245857    -23.63013,-46.704887      -23.7074,-46.5698 
[7110052941, 7110107795]  7100019838        -23.609,-46.6974  -23.657433,-46.744095
[7110052941, 7110107795]  7100021875        -23.609,-46.6974    -23.7729,-46.591366
[7110052941, 7110107795]  0700245857        -23.609,-46.6974       -23.7074,-46.569

To do that first I tried to do something like this:

df["clients_x"] = "[" + "df["clients_x"]" + "]"

However, by doing that will actually add "[ ]" at the beginning and end of each value, but for those rows that already have "[ ]" will duplicate them. The output is this one:

 clients_x                 clients_y              coords_x               coords_y 
[7110001002]                7100019838    -23.63013,-46.704887  -23.657433,-46.744095   
[7110001002]                7100021875    -23.63013,-46.704887    -23.7729,-46.591366   
[7110001002]                0700245857    -23.63013,-46.704887      -23.7074,-46.5698 
[[7110052941, 7110107795]]  7100019838        -23.609,-46.6974  -23.657433,-46.744095
[[7110052941, 7110107795]]  7100021875        -23.609,-46.6974    -23.7729,-46.591366
[[7110052941, 7110107795]]  0700245857        -23.609,-46.6974       -23.7074,-46.569

To avoid that issue I've tried using the following code where basically I want to add "[ ]" at the beginning and at the end of each value in the clients_x column that starts with a digit.

df['clients_x'] = df['clients_x'].mask(df['clients_x'].astype(str).str.startswith(r'^\d'), f'[{df.clients_x}]')

However, the output that this line of code is generating is the same as my original dataframe. If anyone has any idea about how to approach this problem I would really appreciate your help.

Answer 1

Use np.where -

df['clients_x'] = np.where(df['clients_x'].str.startswith('['), df['clients_x'], '[' + df['clients_x'] + ']')

Using df.where -

df['clients_x'].where(df['clients_x'].str.startswith('['), '[' + df['clients_x'] + ']')

Output

0               [7110001002]
1               [7110001002]
2               [7110001002]
3    [7110052941,7110107795]
4    [7110052941,7110107795]
5    [7110052941,7110107795]
Name: clients_x, dtype: object

Answer 2

You need to use where , not mask (see the doc ):

df["clients_x"] = df.clients_x.where(
  df.clients_x.astype(str).str.startswith("["), 
  "[" + df.clients_x + "]"
)