explainable-exceptions 0.0.3

A Python module to explain exceptions in execution-time using a magic cell

explainable-exceptions

Artificial Intelligence is being used in several fields, including computer engineering and programming.

During this work, we pretend to facilitate the process of programming as much as possible adding the option to use a NLP (Natural Language Processing) model to help the user to understand and solve the exceptions that may appear while programming.

As simple as using a magic cell, the user will be able to send the exception to the huggingface chat and obtain a possible solution to solve the exception. In this process, the user must be logged in his huggingchat account. Credentials can be saved to be requested only once.

After receiving the response, the user will be able to request another response if the given one is not helpful enough.

Install

To install explainable_exceptions you must use the following pip command:

pip install explainable_exceptions

How to use

In order to use the magic cell, we need to import it from the core of explainable_exceptions. The cell is called explain.

After that, we can label cells using the magic cell command (%%). We can provide a huggingchat user or we can just use the command (if there are not credentials, then it will ask to generate them).

If an exception is raised inside a magic cell, it will be sent to the huggingchat model to generate a possible solution. This response will be shown by the jupyter standard output.

import random

from explainable_exceptions.core import explain

The following cell must be preceded by:

%%explain <username>

Where <username> must be replaced (if provided) by the username of the huggingchat account that will be used to request information.

random.choice([])

CRITICAL:root:Traceback (most recent call last):
  File "/home/ruescog/explainable-exceptions/explainable_exceptions/core.py", line 78, in explain
    exec(cell, globals(), local_ns)
  File "<string>", line 3, in <module>
  File "/grupoa/config/miniconda3/lib/python3.8/random.py", line 290, in choice
    raise IndexError('Cannot choose from an empty sequence') from None
IndexError: Cannot choose from an empty sequence

WARNING:root:Be careful, the following response has been generated automatically by a Natural Language Processing Model, so the answer may be incorrect or false.

Huggingchat response (online version):

Hello! That error message suggests that you are trying to use the choice() function from the random module on an empty list or sequence. In order to fix the error, ensure that your code creates and assigns a non-empty sequence for random.choices(). Here is an example of how to create a simple randomized sequence using the list constructor:

import random
seq = [1, 2, 3]
random.shuffle(seq)
print("Shuffled sequence:")
for item in seq:
    print(item)

This should randomly shuffle the elements within the sequence and generate output like this:

Shuffled sequence:
2
3
1

Let me know if there’s something else I could assist you with!

---

Do you need another answer? (y/n) n

IndexError: Cannot choose from an empty sequence