src/data/processors.py

"""
Data processing utilities for the leaderboard application.
"""
import pandas as pd
import numpy as np

def apply_value_formatting(value, is_numeric=True):
    """
    Apply formatting to a value based on its properties
    
    Args:
        value: The value to format
        is_numeric (bool): Whether the value is numeric
        
    Returns:
        dict: Dictionary with formatting information
    """
    if not is_numeric or value == '-':
        return {'value': value, 'class': ''}
    
    numeric_value = float(value)
    if numeric_value > 0:
        return {'value': value, 'class': 'positive-value'}
    elif numeric_value < 0:
        return {'value': value, 'class': 'negative-value'}
    else:
        return {'value': value, 'class': ''}

def get_model_type_style(model_type):
    """
    Get styling for different model types
    
    Args:
        model_type (str): The model type
        
    Returns:
        dict: Dictionary with styling information
    """
    if model_type == "Open Source":
        return {'color': '#4ade80'}  # Brighter green
    elif model_type == "Open Weights":
        return {'color': '#93c5fd'}  # Brighter blue
    elif model_type == "Closed Source":
        return {'color': '#cbd5e1'}  # Lighter gray
    else:
        return {'color': ''}

def get_rank_style(rank):
    """
    Get styling for different ranks
    
    Args:
        rank (str): The rank
        
    Returns:
        dict: Dictionary with styling information
    """
    if "🥇" in str(rank):
        return {'color': 'gold', 'font-weight': '700', 'font-size': '16px'}
    elif "🥈" in str(rank):
        return {'color': 'silver', 'font-weight': '700', 'font-size': '16px'}
    elif "🥉" in str(rank):
        return {'color': '#cd7f32', 'font-weight': '700', 'font-size': '16px'}
    else:
        return {}

def calculate_task_statistics(metric_data):
    """
    Calculate statistics for each task
    
    Args:
        metric_data (dict): Dictionary containing the metric data
        
    Returns:
        dict: Dictionary with task statistics
    """
    stats = {}
    for task, models in metric_data.items():
        values = list(models.values())
        stats[task] = {
            'mean': np.mean(values),
            'median': np.median(values),
            'min': min(values),
            'max': max(values),
            'std': np.std(values)
        }
    return stats