Split dict value tuples into df headers and contents in python3

Question

I have a dictionary that looks like this:

{'Customer A': {'key1':'value 1',
                'fees': [[0, 0.26], [5, 0.24], [10, 0.22], [25, 0.2],
                         [50, 0.18], [100, 0.16], [250, 0.14],
                         [500, 0.12], [1000, 0.1]],
                'key3':'value3'},
{'Customer B':...

How do I get a 'fees' df that looks like this:

             0     5    10    25    50   100   250   500   1000  
Customer A  0.26  0.24  0.22  0.2  0.18  0.16  0.14  0.12  0.1
Customer B

Customers become the Index and first part of value tuple becomes the header

Still learning here, so all help is appreciated. Unfortunately this doesn't feel like basic dictionary/list/df operations...

score 3 · Accepted Answer · answered Mar 02 '18 at 15:25

Use dict comprehension with DataFrame.from_dict, last sort columns by sort_index:

d = {'Customer A': {'key1':'value 1',
                'fees': [[0, 0.26], [5, 0.24], [10, 0.22], [25, 0.2],
                         [50, 0.18], [100, 0.16], [250, 0.14],
                         [500, 0.12], [1000, 0.1]],
                'key3':'value3'}, 'Customer B': {'key1':'value 1',
                'fees': [[0, 0.26], [5, 0.24], [10, 0.22], [25, 0.2],
                         [50, 0.18], [100, 0.16], [250, 0.14],
                         [500, 0.12], [1000, 0.1]],
                'key3':'value3'}}


df = pd.DataFrame.from_dict({k: dict(v['fees']) for k, v in d.items()}, orient='index')
print (df)
            0     50    100   5     1000  25    10    500   250 
Customer A  0.26  0.18  0.16  0.24   0.1   0.2  0.22  0.12  0.14
Customer B  0.26  0.18  0.16  0.24   0.1   0.2  0.22  0.12  0.14

df = df.sort_index(axis=1)
print (df)
            0     5     10    25    50    100   250   500   1000
Customer A  0.26  0.24  0.22   0.2  0.18  0.16  0.14  0.12   0.1
Customer B  0.26  0.24  0.22   0.2  0.18  0.16  0.14  0.12   0.1

How can I avoid an error if a customer doesn't have a 'fees' value? In which case the comprehension iteration would just go onto the next customer — ACF, Mar 02 '18 at 16:14
@ACF Can you create new question? Now I am offline, on phone only. Thanks. — jezrael, Mar 02 '18 at 16:17
@jazrael New question is here: [link](https://stackoverflow.com/q/49077197/9242601) — ACF, Mar 02 '18 at 20:51

Split dict value tuples into df headers and contents in python3

1 Answers1

Linked