使用'.format（）'的模拟在Python中分配变量

Question

I have a dataset with several columns of which I want to convert from categorical to numerical dtype.

I have defined the following function:

def create_num_column(df,column):
    df[str(column)]=df[str(column)].astype('category')
    df[str(column)+'_cat']=df[str(column)].cat.codes

    dictionary{}.format(str(column)+'_cat')=dict(zip(df[column],df[str(column)+'_cat']))
    return dictionary{}.format(str(column)+'_cat')

I also want this function to create a dictionary so I will be able to understand what label was assigned to what value.

The problem here arises with naming of a dictionary as I have tried to use '.format()' and it has shown syntax error.

dictionary{}.format(str(column)+'_cat')=dict(zip(df.column,df[str(column)+'_cat']))

>SyntaxError: invalid syntax

I understand that this particular method works with strings but is there any way to 'automate' the naming of variable?

Thanks in advance.

Answer 1

You cannot "dynamically" name a variable, and there is no need to do so. Simply return the dictionary:

def create_num_column(df,column):
    df[str(column)]=df[str(column)].astype('category')
    df[str(column)+'_cat']=df[str(column)].cat.codes

    return dict(zip(df.column,df[str(column)+'_cat']))

(There is a way to create variables or more generally objects with dynamic names in Python, but that's only very rarely useful.)