Population

pangadfs.populate

PopulateDefault()

Bases: PopulateBase

Source code in pangadfs/base.py

def __init__(self):
    logging.getLogger(__name__).addHandler(logging.NullHandler())

populate(*, pospool, posmap, population_size, probcol='prob', **kwargs)

Creates individuals in population

Parameters:

Name	Type	Description	Default
pospool	Dict[str, DataFrame]	pool split into positions	required
posmap	Dict[str, int]	positions & accompanying roster slots	required
population_size	int	number of individuals to create	required
probcol	str	the dataframe column with probabilities	'prob'
**kwargs		keyword arguments	{}

Returns:

Type	Description
ndarray	ndarray of size (population_size, sum(posmap.values()))

Source code in pangadfs/populate.py

def populate(self,
             *, 
             pospool, 
             posmap: Dict[str, int], 
             population_size: int, 
             probcol: str='prob',
             **kwargs) -> np.ndarray:
    """Creates individuals in population

    Args:
        pospool (Dict[str, DataFrame]): pool split into positions
        posmap (Dict[str, int]): positions & accompanying roster slots
        population_size (int): number of individuals to create
        probcol (str): the dataframe column with probabilities
        **kwargs: keyword arguments

    Returns:
        ndarray of size (population_size, sum(posmap.values()))

    """
    pos_samples = {
        pos: multidimensional_shifting(pospool[pos].index, population_size, n, pospool[pos][probcol])
        for pos, n in posmap.items()
    }

    # Handle FLEX separately to maintain original behavior
    if 'FLEX' in posmap:
        # concatenate non-FLEX positions into single row
        non_flex_samples = [pos_samples[pos] for pos in posmap if pos != 'FLEX']
        if non_flex_samples:
            pop = np.concatenate(non_flex_samples, axis=1)
        else:
            pop = np.empty((population_size, 0), dtype=int)

        # For FLEX, take the correct number of positions as specified in posmap
        flex_count = posmap['FLEX']
        flex_sample = pos_samples['FLEX'][:, :flex_count]  # Take correct number of FLEX positions

        return np.column_stack((pop, flex_sample))
    else:
        # No FLEX position, concatenate all
        return np.concatenate([pos_samples[pos] for pos in posmap], axis=1)