src/populate.py

"""
Data population functions for MUSEval Leaderboard
"""

import pandas as pd
from typing import Dict, List, Any, Optional
from .load_results import load_results_with_metadata, create_overall_table

def get_leaderboard_df(results_path: str, requests_path: str, eval_cols: List[str], benchmark_cols: List[str]) -> pd.DataFrame:
    """Get leaderboard dataframe"""
    # Use our existing load_results function
    results = load_results_with_metadata()
    if not results:
        return pd.DataFrame()
    
    return create_overall_table()

def get_model_info_df(results_path: str, requests_path: str) -> pd.DataFrame:
    """Get model information dataframe"""
    results = load_results_with_metadata()
    if not results:
        return pd.DataFrame()
    
    # Extract unique model information
    model_info = {}
    for result in results:
        model = result["model"]
        if model not in model_info:
            model_info[model] = {
                "model": model,
                "organization": result["submitter"],
                "submission_date": result["submission_date"],
                "task": result.get("task", ""),
                "dataset_version": result.get("dataset_version", ""),
                "paper_url": result.get("paper_url", ""),
                "code_url": result.get("code_url", ""),
                "model_type": "Foundation Model",  # Default
                "testdata_leakage": "No"  # Default
            }
    
    return pd.DataFrame(list(model_info.values()))

def get_merged_df(leaderboard_df: pd.DataFrame, model_info_df: pd.DataFrame) -> pd.DataFrame:
    """Merge leaderboard and model info dataframes"""
    if leaderboard_df.empty or model_info_df.empty:
        return leaderboard_df
    
    # Merge on model name
    merged = pd.merge(leaderboard_df, model_info_df, on="model", how="left")
    
    # Add rank column
    if 'MAE' in merged.columns:
        merged['Rank'] = merged['MAE'].rank(method='min').astype(int)
        # Move Rank to front
        cols = ['Rank'] + [col for col in merged.columns if col != 'Rank']
        merged = merged[cols]
    
    return merged

def get_evaluation_queue_df(requests_path: str, eval_cols: List[str]) -> tuple:
    """Get evaluation queue dataframes"""
    # Return empty dataframes for now
    return pd.DataFrame(), pd.DataFrame(), pd.DataFrame()