LaloLoop / most_common_path

Most common path challenge python implementation

Most common path challenge

Objective

To find the most common path for a user status change from different statuses in a given transaction log.

Restrictions

To obtain a path, the user should have 3 different states on his journey.

Example

If a user starts on the state A, then go to state B, then move forward to state C and finalize with state D. Then the user creates two different paths:

1. A->B->C
2. B->C->D

Input format: [<userId>,<stateId>]

How to run this code

Main code

The main code has a sample test case hardcoded as an example of the functionality. It can be executed with the following command:

python main.py

Unit tests

You can find test cases within the test_most_common_path.py file. The easiest way to run this is with the following command:

python -m unittest

Improvements

There are some edge cases that are not considered:

1. What happens if a user can somehow edit the path and add a new state in between (not an append only log)?
2. What happens if the user id or state id change to something other than number and letters? The current implementation may not be the best option memory wise.

Another implementation

Another implementation could be achieved with a different data structure, like a weighted graph and the traversal on the graph could help find out the different common paths.