feat: Implement tournament ranking algorithm for action prioritization #15
Conversation
- Replaced previous qualitative prioritization method with tournament ranking - Updated output generation to use tournament ranking for both adaptation and mitigation actions - Removed detailed explanations and replaced with generic ranking explanation to maintain same data structure - Maintained existing output file structure and naming conventions
![korbit-ai[bot]](https://avatars.githubusercontent.com/in/322216?s=60&v=4){kind=small}
There was a problem hiding this comment.
Review by Korbit AI
Korbit automatically attempts to detect when you fix issues in new commits.
| Category | Issue | Fix Detected |
|---|---|---|
 |
Incomplete Tournament Error Handling ▹ view | |
 |
Missing ML Comparison Result Caching ▹ view |
Files scanned
| File Path | Reviewed |
|---|---|
| prioritizer/prioritizer.py | ✅ |
Explore our documentation to understand the languages and file types we support and the files we ignore.
Need a new review? Comment
/korbit-reviewon this PR and I'll review your latest changes.Korbit Guide: Usage and Customization
Interacting with Korbit
- You can manually ask Korbit to review your PR using the
/korbit-reviewcommand in a comment at the root of your PR.- You can ask Korbit to generate a new PR description using the
/korbit-generate-pr-descriptioncommand in any comment on your PR.- Too many Korbit comments? I can resolve all my comment threads if you use the
/korbit-resolvecommand in any comment on your PR.- Chat with Korbit on issues we post by tagging @korbit-ai in your reply.
- Help train Korbit to improve your reviews by giving a 👍 or 👎 on the comments Korbit posts.
Customizing Korbit
- Check out our docs on how you can make Korbit work best for you and your team.
- Customize Korbit for your organization through the Korbit Console.
Current Korbit Configuration
General Settings
Setting Value Review Schedule Automatic excluding drafts Max Issue Count 10 Automatic PR Descriptions ✅ Issue Categories
Category Enabled Documentation ✅ Logging ✅ Error Handling ✅ Readability ✅ Design ✅ Performance ✅ Security ✅ Functionality ✅ Feedback and Support
Note
Korbit Pro is free for open source projects 🎉
Looking to add Korbit to your team? Get started with a free 2 week trial here
| while remaining and current_rank <= 20: | ||
| print(f"\n--- Running bracket for rank #{current_rank} with {len(remaining)} actions ---") | ||
| winner, losers = single_elimination_bracket(remaining, city) | ||
|
|
||
| if not winner: | ||
| #TODO is there a normal thing that this can happen ?or should this be error | ||
| print("No winner found, breaking") | ||
| break |
There was a problem hiding this comment.
Incomplete Tournament Error Handling 
Tell me more
What is the issue?
The tournament_ranking function has an unresolved TODO about error handling when no winner is found, leaving uncertainty about proper error handling.
Why this matters
Without proper error handling, the tournament might terminate prematurely or produce incomplete rankings without clear indication of failure.
Suggested change ∙ Feature Preview
Implement proper error handling for the no-winner case:
while remaining and current_rank <= 20:
winner, losers = single_elimination_bracket(remaining, city)
if not winner:
print(f"Error: Tournament failed at rank {current_rank}")
if not full_ranking:
raise RuntimeError("Tournament failed without producing any rankings")
return full_ranking # Return partial results if we have any
💬 Chat with Korbit by mentioning @korbit-ai.
|
|
||
|
|
||
| # Use your ML model to compare | ||
| result = ML_compare(actionA, actionB, city) |
There was a problem hiding this comment.
Missing ML Comparison Result Caching 
Tell me more
What is the issue?
No caching of ML comparison results between the same action pairs.
Why this matters
Redundant ML model calls for previously compared action pairs waste computational resources and increase API costs.
Suggested change ∙ Feature Preview
Implement a comparison cache to store and reuse ML comparison results:
comparison_cache = {}
def cached_ML_compare(actionA, actionB, city):
pair_key = tuple(sorted([actionA['ActionID'], actionB['ActionID']]))
if pair_key not in comparison_cache:
comparison_cache[pair_key] = ML_compare(actionA, actionB, city)
return comparison_cache[pair_key]💬 Chat with Korbit by mentioning @korbit-ai.
| @@ -329,31 +328,211 @@ def filter_actions_by_biome(actions, city): | |||
| ] | |||
|
|
|||
|
|
|||
| def ML_compare(actionA, actionB, city): | |||
There was a problem hiding this comment.
why do you have this wrapper? It does essentially just return the result of my implemented ml_compare so you could use that directly here
| if wildcard: | ||
| winners.append(wildcard) | ||
| print(f" Wildcard {wildcard.get('ActionName', 'Unknown')} automatically advances") | ||
|
|
There was a problem hiding this comment.
I think its better to use the action_id instead of the name? The id is easier for identifying an action compared to the name
| # If exactly one winner, we found the bracket winner | ||
| if len(winners) == 1: | ||
| print(f" Final winner of bracket: {winners[0].get('ActionName', 'Unknown')}") | ||
| return winners[0], losers |
There was a problem hiding this comment.
same here? action_id vs action_name
| print(" No winner found, breaking") | ||
| break # no more participants | ||
|
|
||
| print(f" Rank #{rank}: {winner.get('ActionName', 'Unknown')}") |
There was a problem hiding this comment.
We could use ActionName as well I guess. Just feeling action_id is unique and the correct identifier.
| wildcard = None | ||
| if len(actions) % 2 == 1: | ||
| wildcard = actions.pop() | ||
| print(f" Odd number of actions, wildcard: {wildcard.get('ActionName', 'Unknown')}") |
| print(f" Round complete. {len(winners)} winners advancing to next round") | ||
|
|
||
| # If exactly one winner, we found the bracket winner | ||
| if len(winners) == 1: |
There was a problem hiding this comment.
I am not sure I understand the tournament ranking completely.
In what case would len(winners) == 1?
We pass in a list of x actions.
Then we do a 1 against 2, 3 against 4, 5 against 6 and so on comparison and append all the winners to the winners list.
So we have as many winners as we have matches no?
Description by Korbit AI
What change is being made?
Implement a tournament ranking algorithm for action prioritization, replacing previous quantitative score-based rankings across various city-specific adaptation and mitigation JSON datasets.
Why are these changes being made?
This modification aims to standardize the ranking process of adaptation and mitigation actions using a consistent tournament ranking approach, improving the prioritization mechanism by providing a uniform rationale ("Ranked by tournament ranking algorithm") across cities such as Camaçari, Corumbá, Caxias do Sul, and Maranguape. The change is intended to enhance clarity and objectivity in the decision-making process for stakeholders reviewing climate action plans.