replace() takes no keyword arguments, in for loop

Question

I'm trying to convert multiple columns of dollar amount into float, and wrote the following code

    for column in wo.columns[14:21]:
        column = (column.replace( '[\$,)]','', regex=True )
                   .replace( '[(]','-',   regex=True ).replace('#NAME?','NaN', regex=True).astype(float))
    return column


And this is the error i get:
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-276-f7c13cb3d0af> in <module>
      1 for column in wo.columns[14:19]:
----> 2     column = (column.replace( '[\$,)]','', regex=True )
      3                .replace( '[(]','-',   regex=True ).replace('#NAME?','NaN', regex=True).astype(float))
      4 return column
      5 

TypeError: replace() takes no keyword arguments

What might be wrong? wo is the name of the dataframe, the library i used to load the dataframe was Pandas, and when i used the code on other individual columns it worked fine, just when i used a for loop it return an error.

Solution 1

It is giving an error because it doesn't accept keyword arguments; in this case regex=True. As per the documentation, it doesn't exist. Also, .replace is for strings only, so the type of data here should be str. On the other hand, should you want to use regex, I'd recommend re.sub

Solution 2

Try to delete the 'regex = True' will be helpful from my case.

# df_num_key_split[name][var].replace(to_replace="[\[|\]|']|\s+|\.|\-", value='', regex=True)

After I did this, it worked.

df_num_key_split[name][var].replace("[\[|\]|']|\s+|\.|\-", '')

Comments

[Michael Butscher]

If you use an external library add it to the tags of the question. And show definition of "wo".

[fabianegli]

If you use pandas, you need to apply the replace to the str of the DataFrame or Series like so: column.str.replace. The str part has to be added every time before the replace function.

[fabianegli]

Also, pandas regex=True is the default and does not need to be specified.