Remove leading zeroes in pandas column but only for numeric-CodePudding

My pandas dataframe looks as follows:

Now I want to remove the leading zeroes, if the string is only numerical. Any suggestions? So outcome should be:

CodePudding user response：

You can use a regex with str.replace for this:

df['col2'] = df['col2'].str.replace(r'^0 (?!.*\D)', '', regex=True)

output:

   col1        col2
0     1  ABC8392akl
1     2        1523
2     3    000ABC58

regex:

^0        # match leading zeros
(?!.*\D)  # only if not followed at some point by a non digit character

suggested by @timgeb

df['col2'] = df['col2'].str.replace(r'^0 (\d*)$', r'\1', regex=True)

regex:

^0        # match leading zeros
(\d*)     # capture other digits (if any)
$         # match end of string

replace with captured digits (\1)

CodePudding user response：

Use

where = (df['col2'].str.isdigit(), 'col2')
df.loc[where] = df.loc[where].str.lstrip('0')

CodePudding user response：

Is there a problem with this, or you want to keep the numbers as string?

df['col2'] = pd.to_numeric(df['col2'] ,errors='ignore')