Statistical comparison between a candidate and baseline pipeline.
| Name | Type | Description | Notes |
|---|---|---|---|
| candidate_retriever_id | str | ID of the candidate pipeline. | |
| ndcg_delta | Dict[str, float] | Change in NDCG at each K (positive = candidate better). | |
| recall_delta | Dict[str, float] | Change in recall at each K (positive = candidate better). | |
| latency_delta_ms | float | Change in mean latency (positive = candidate slower). | |
| p_value | float | Statistical significance of the difference (paired t-test). | [optional] |
| confidence_interval | List[object] | 95% confidence interval for NDCG@10 delta. | [optional] |
| taxonomy_wins | List[str] | Taxonomy nodes where candidate significantly outperforms. | [optional] |
| taxonomy_losses | List[str] | Taxonomy nodes where candidate significantly underperforms. | [optional] |
from mixpeek.models.pipeline_comparison import PipelineComparison
# TODO update the JSON string below
json = "{}"
# create an instance of PipelineComparison from a JSON string
pipeline_comparison_instance = PipelineComparison.from_json(json)
# print the JSON string representation of the object
print(PipelineComparison.to_json())
# convert the object into a dict
pipeline_comparison_dict = pipeline_comparison_instance.to_dict()
# create an instance of PipelineComparison from a dict
pipeline_comparison_from_dict = PipelineComparison.from_dict(pipeline_comparison_dict)