Filter dictionary with values of different types

Question

I am trying to filter dictionary by value (dictionary with different data types), but got the error:

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

I want to get all records corresponding to 'YALE' value in my dictionary.

This is my code:

dataset = {
    'timeseires': array([[
        [ -5.653222,   7.39066 ,  20.651941, 4.07861 ,-11.752331, -34.611312],
        [ -5.653222,   7.39066 ,  20.651941, 4.07861 ,-11.752331, -34.611312]
    ]]),
    'site': array(['YALE', 'KKI'], dtype='<U8')
}

dataset = data.tolist()
 
def filter(pairs):
    key, value = pairs
    filter_key = 'site'
    if key == filter_key and value == 'YALE':
        return True
    else:
        return False
         
final_dic = dict(filter(filter, dataset.items()))
print(final_dic)

The expected output:

> dataset = {
>         'timeseires': array([[
>             [ -5.653222,   7.39066 ,  20.651941, 4.07861 ,-11.752331, -34.611312]
>         ]]),
>         'site': array(['KKI'], dtype='<U8')
>     }

Answer 1

From the output given, it seems like your task is to extract the 'timeseries' value corresponding to the index of the 'YALE'.

This code should do the trick:

def filter_dictionary_by_value(dataset, value):
    index_value = np.where(dataset['site'] == value)[0]
    filtered_timeseries = dataset['timeseries'][index_value]
    

    filtered_dict = {
        'timeseries': filtered_timeseries,
        'site': np.array([value], dtype='<U8')
    }

If this wasn't what you were expecting please be more specific in the question.

Answer 2

Considering your expected output, you could do something like this using dict comprehension:

import numpy as np

filtered_indices = np.where(dataset["site"] == "YALE")[0].tolist()
dataset = {k: np.delete(v, filtered_indices) for k, v in dataset.items()}

{'timeseires': array([  7.39066 ,  20.651941,   4.07861 , -11.752331, -34.611312,
        -5.653222,   7.39066 ,  20.651941,   4.07861 , -11.752331,
       -34.611312]), 'site': array(['KKI'], dtype='<U8')}