Is this a better way than pd.get_dummies ?

[Programmer-RD-AI]

def object_to_int(data,column): index = -1 all_info = [] info_dict = {} for info in data[column]: if info not in info_dict: index = index + 1 info_dict[info] = index for info in data[column]: all_info.append(info_dict[info]) return (index,all_info,info_dict)

Is the above way better than pd.get_dummies?

[mrdbourke]

Hey there,

It depends what you're trying to do.

To me, the pandas get_dummies method looks cleaner than the above (if you're just using the method).

Have you test the function versus the get_dummies method? Does it achieve what you're after?

[Programmer-RD-AI]

I have tried OneHotEncoder in sklearn I use it most of the time, but I was just curious to know what is better the custom function or pd.get_dummies()

Thanks

[mrdbourke]

You should generally favour pre-built methods and functions from libraries such as pandas and Scikit-Learn over any custom-built functions.

Functions that exist in libraries have been tested by many people.

[Programmer-RD-AI]

OK Thank you so much