Can't get the size of numpy.ndarray

Question

I have a dataframe as follows:

  version  count region      listing
2      v2      2    CAN     [7, 8]
2      v3      3    CAN  [7, 8, 9]

I want to extract listing list for each row and get the length. So I did the following:

group_v2_list = group[group['version'] == 'v2']['listing'].values

and I get output as [list([7, 8])] . Here the type of listing column is numpy.ndarray which I get after using type(group_v2_list) .

Now I want to get the number of elements in this group_v2_list but I am unable to get it.

I tried len(group_v2_list) and group_v2_list.size but both are giving me 1 . I want to get the number of elements which should be 2 as 7, 8 .

How can I get that?

Answer 1

You do not need to access the numpy representation for this.

One way is to use .loc accessor to extract the series and find the length of the first element:

df = pd.DataFrame({'version': ['v2', 'v3'],
                   'count': [2, 3],
                   'region': ['CAN', 'CAN'],
                   'listing': [[7, 8], [7, 8, 9]]})

df_v2_list = df.loc[df['version'] == 'v2', 'listing']

res_v2 = len(df_v2_list[0])
# 2

If there are multiple elements in your filtered data, you can retrieve a list of their lengths by using pd.Series.map(len) :

df_v_all_list = df.loc[df['version'].str.startswith('v'), 'listing']

res_all = df_v_all_list.map(len).tolist()
# [2, 3]