relari-ai
Aggregate classification metrics
@ellipsis-dev can you give this a try?
Create a new folder classification under metrics, then a file called classification_metrics.py. In the file, you need to create a class ClassificationAccuracy(Metric):
In the call function, calculate the match between the predicted_class and the ground_truth_classes. If there's a match then return {"classification_correctness": 1.0 or 0.0"}
In the aggregate function, make sure to output, accuracy precision recall f1. Take into account both multi-class and binary-class situations
I can't make an implementation plan because the feature request lacks specific details about what changes are required or how the aggregation should be performed. Without these details, it's not possible to create an accurate implementation plan.
![ellipsis-dev[bot] avatar](https://avatars.githubusercontent.com/in/64358?v=4 "Published by a pub.dev verified publisher"){kind=small}
@ellipsis-dev How about you forget about the aggregation. Just implement the call function that checks if the predicted class matches the ground_truth_class
Sorry, Ellipsis encountered a problem while generating a pull request. Our team has been alerted and is investigating. (wflow_BNdnWAxlKTfF8qNo) :robot:
Apologies, there’s a bug with Ellipsis where it’s not properly receiving your comments, only the issue body (hence its confusion both times) - i’m fixing now
Sorry, Ellipsis encountered a problem while generating a pull request. Our team has been alerted and is investigating. (wflow_IcdOq4eQYJHyqWZo) :robot: