How to dynamically extract data from multiple Python dict entries

user_2102292 0
Viewed 199

Let's say I have a dict like this:

my_dict = {something: 'blabla', result: 'something', value_0_0: 'apple', value_0_1: 'ball', value_1_0: 'banana', value_1_1: 'car', value_2_0: 'orange', value_2_1: 'toy'}

The dict may have other key-value entries with other names like: result: 'something'. I just want to filter those keys with the following structure:

value_X_Y

Desired output:

0values = "apple ; banana ; orange"

1values = "ball ; car; toy"

If the size of the dictionary was fixed, you could do something like this:

for x in my_dict:
    if(x == 'value_0_0'):
        #do something
    if(x == 'value_0_1'):
        #do something
    if(x == 'value_1_0'):
        #do something
    if(x == 'value_1_1'):
        #do something
    if(x == 'value_2_0'):
        #do something
    if(x == 'value_2_1'):
        #do something

But since the dictionary do not have a fixed size and can have 3 values or 50 (for example) value_50_0, value_50_1, I would like to process the data dynamically.

How could I get something like that?

Any insights will be appreciated.

python
5 Answers

The following example uses a regex to identify the keys starting with "value_" and captures the last number to use it to create a new dictionary with an array and the related values:

import re

my_dict = {'something': 'blabla', 'result': 'something','value_0_0': 'apple', 'value_0_1': 'ball', 'value_1_0': 'banana', 'value_1_1': 'car', 'value_2_0': 'orange', 'value_2_1': 'toy'}
result = {}

for key, value in my_dict.items():
    # Get last number from the key to be used as the key of the result array
    m = re.match(r'^value\_\d+\_(\d+$)', key)
    if m is not None:
        result_key = m.groups()[0] + 'values'
        if not result_key in result:
            result[result_key] = []
        
        result[result_key].append(value)

print(result)

Result:

{'0values': ['apple', 'banana', 'orange'], '1values': ['ball', 'car', 'toy']}
user_26730290 
my_dict = {'value_0_0': 'apple', 'value_0_1': 'ball', 'value_1_0': 'banana', 'value_1_1': 'car','value_2_0': 'orange', 'value_2_1': 'toy'}
fruits,acces=[],[]
for key,value in my_dict.items():
    if 0 == int(key.split('_')[-1]):
        fruits.append(value)
    if 1 == int(key.split('_')[-1]):
        acces.append(value)

fruits and acces will be a list with respective data

fruits = ['apple', 'banana', 'orange']
acces = ['ball', 'car', 'toy']
user_159489870

If you're just looking to iterate through the dictionary, you can use the dict.keys(), dict.values(), or dict.items() methods. You could then create a new dictionary mapping the various values ("0values", "1values", etc.) to the appropriate string values.

Something like this, perhaps:

import re

values = {}

# Create a regular expression to match strings in the format 'value_X_Y'
regex = re.compile(r"^value_\d+_(\d+)$")

for key, val in my_dict.items():
    # Try to match 'Y' in 'value_X_Y', or else skip the key-value pair
    keynum = regex.search(key)
    if keynum is None:
        continue

    # Update dictionary with formatted key and value 
    value_key = f"{keynum.group(1)}values"
    values.setdefault(value_key, [])
    values[value_key].append(val)

Here, the dict.setdefault() method will create the key value_key and set it to [], provided that value_key does not already exist in the dictionary. This would be the same as checking for the key in the dictionary 'manually':

if value_key not in values:
    values[value_key] = []

With the dictionary provided, some sample usage:

>>> import re
>>> my_dict = {
    "value_0_0": "apple",
    "value_0_1": "ball",
    "value_1_0": "banana",
    "value_1_1": "car",
    "value_2_0": "orange",
    "value_2_1": "toy"
}
>>> values = {}
>>> regex = re.compile(r"^value_\d+_(\d+)$")
>>> for key, val in my_dict.items():
        keynum = regex.search(key)
        if keynum is None:
            continue
        value_key = f"{keynum.group(1)}values"
        values.setdefault(value_key, [])
        values[value_key].append(val)

>>> values
{'0values': ['apple', 'banana', 'orange'], '1values': ['ball', 'car', 'toy']}

Then with some formatting, it's not too hard to get the desired output:

>>> for key, val in values.items():
        print(f"{key}: {'; '.join(val)}")

    
0values: apple; banana; orange
1values: ball; car; toy
user_137463890

You can use the split method here on keys of the dictionary to extract the numbers and then process the information.

my_dict = {"value_0_0": 'apple', "value_0_1": 'ball', "value_1_0": 'banana', "value_1_1": 'car', "value_2_0": 'orange', "value_2_1": 'toy'}

results = {}

for key, value in my_dict.items():
    _, _, y = key.split("_")

    if y not in results:
        results[y] = []

    results[y].append(value)

for k, v in results.items():
    print(k+"values", "=", "; ".join(v))
user_125660540

With using re for matching keys and defaultdict to avoid checking if an element exists in the dict and some short circuiting with and you can do somthing like this:

import re
from collections import defaultdict

res = defaultdict(list)

for key, val in my_dict.items(): 
    (r := re.findall('value_\d+_(\d+)', key)) and res[f'{r[0]}values'].append(val)

res = dict(res)

You can leave out the last line if you're okay with the defaultdict as an outcome.

user_142529550
Related