Is there a difference between series.tolist() and list in pandas

Question

I defined a function z. It works when I pass a list, but when I pass a series(even after converting to list) it returns wrong answers. My input argument to the function z has to be a series. How to resolve this?

list1 = [np.nan, 14975, 98121] series1 = pd.Series([np.nan,14975,98121]) z(series1.tolist()) ['0', '0', '0'] z(list1) ['0', '1', '98121']

My z function is,

def z(each): zipcode_list = [] for i in each: try: if zipcodes.is_real(str(i)): zip_code = str(i) else: zip_code = str(1) except: zip_code = str(0) zipcode_list.append(zip_code) return zipcodes

they all return zeros to me, there's no function is_real for lists, so they all should evaluate to zip_code = str(0) — Yuca
– Yuca, Commented Aug 1, 2019 at 21:50
Did you try importing 'zipcodes' module, without that it will print the exception always. — Anitha
– Anitha, Commented Aug 1, 2019 at 21:54
there's no zipcodes module on the anaconda base envirronment, if there's one import called zipcodes in your code then rename your variable zipcodes inside you z function — Yuca
– Yuca, Commented Aug 1, 2019 at 21:56

OrionTheHunter · Accepted Answer · 2019-08-01 22:21:19Z

Although pandas does correctly return a list (see https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Series.tolist.html) they take the safe route and turn the values in the Series to floats and not ints. Thus when you try to get the zipcode of a float, it errors.

You can see this by running the following:

import pandas as pd import numpy as np import zipcodes list1 = [np.nan, 14975, 98121] series1 = pd.Series([np.nan,14975,98121]) def z(each): zipcode_list = [] for i in each: print(i, type(i)) try: if zipcodes.is_real(str(i)): zip_code = str(i) else: zip_code = str(1) except Exception: zip_code = str(0) zipcode_list.append(zip_code) return zipcode_list print(z(series1.tolist())) print(z(list1))

Output:

nan <class 'float'> 14975.0 <class 'float'> import pandas as pd 98121.0 <class 'float'> ['0', '0', '0'] nan <class 'float'> 14975 <class 'int'> 98121 <class 'int'> ['0', '1', '98121']

Changing the code to convert the list to ints before passing it into z will fix your problem. See:

import pandas as pd import numpy as np import zipcodes list1 = [np.nan, 14975, 98121] series1 = pd.Series([np.nan,14975,98121]) def z(each): zipcode_list = [] for i in each: try: if zipcodes.is_real(str(int(i))): zip_code = str(int(i)) else: zip_code = str(1) except Exception: zip_code = str(0) zipcode_list.append(zip_code) return zipcode_list print(z(series1.tolist())) # ['0', '1', '98121'] print(z(list1)) # ['0', '1', '98121']

Thanks, but the 'zipcode' is a module that has is_real() which returns true or false. Without that module it will always print the exception.

Collectives™ on Stack Overflow

Is there a difference between series.tolist() and list in pandas

1 Answer 1

1 Comment

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Related