wehub-resource-sync
34 lines
1022 B
Python
34 lines
1022 B
Python
"""Evaluation framework for code-review-graph.
|
|
|
|
Provides scoring metrics (token efficiency, MRR, precision/recall),
|
|
benchmark runners, and report generators for benchmarking graph-based code reviews.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
from .reporter import generate_full_report, generate_markdown_report, generate_readme_tables
|
|
from .scorer import compute_mrr, compute_precision_recall, compute_token_efficiency
|
|
|
|
|
|
def __getattr__(name: str):
|
|
"""Lazy-import runner functions (require pyyaml)."""
|
|
_runner_names = {"load_all_configs", "load_config", "run_eval", "write_csv"}
|
|
if name in _runner_names:
|
|
from . import runner
|
|
return getattr(runner, name)
|
|
raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
|
|
|
|
|
|
__all__ = [
|
|
"compute_mrr",
|
|
"compute_precision_recall",
|
|
"compute_token_efficiency",
|
|
"generate_full_report",
|
|
"generate_markdown_report",
|
|
"generate_readme_tables",
|
|
"load_all_configs",
|
|
"load_config",
|
|
"run_eval",
|
|
"write_csv",
|
|
]
|