penaltyblog.models.BivariatePoissonGoalModel#

class penaltyblog.models.BivariatePoissonGoalModel(goals_home: Sequence[int] | ndarray[Any, dtype[int64]] | Series, goals_away: Sequence[int] | ndarray[Any, dtype[int64]] | Series, teams_home: Sequence[str] | ndarray[Any, dtype[str_]] | Series, teams_away: Sequence[str] | ndarray[Any, dtype[str_]] | Series, weights: float | Sequence[float] | ndarray[Any, dtype[float64]] | Series | None = None)[source]#

Bases: BaseGoalsModel

Karlis & Ntzoufras Bivariate Poisson for soccer, with:

X = W1 + W3 Y = W2 + W3

where W1, W2, W3 ~ independent Poisson(lambda1, lambda2, lambda3).

__init__(goals_home: Sequence[int] | ndarray[Any, dtype[int64]] | Series, goals_away: Sequence[int] | ndarray[Any, dtype[int64]] | Series, teams_home: Sequence[str] | ndarray[Any, dtype[str_]] | Series, teams_away: Sequence[str] | ndarray[Any, dtype[str_]] | Series, weights: float | Sequence[float] | ndarray[Any, dtype[float64]] | Series | None = None)[source]#

Initialises the BivariatePoissonGoalModel class.

Parameters:

  • goals_home (array_like) – The number of goals scored by the home team

  • goals_away (array_like) – The number of goals scored by the away team

  • teams_home (array_like) – The names of the home teams

  • teams_away (array_like) – The names of the away teams

  • weights (array_like, optional) – The weights of the matches, by default None

Methods

__init__(goals_home, goals_away, teams_home, ...)

Initialises the BivariatePoissonGoalModel class.

fit()

Fits the Bivariate Poisson model to the data.

get_params()

Return the fitted parameters in a dictionary.

load(filepath)

Loads a model from a file.

predict(home_team, away_team[, max_goals])

Predicts the probability of each scoreline for a given home and away team.

save(filepath)

Saves the model to a file using pickle.

Attributes

params

Property to retrieve the fitted model parameters.
fit()[source]#

Fits the Bivariate Poisson model to the data.

get_params() → Dict[str, float][source]#

Return the fitted parameters in a dictionary.

classmethod load(filepath: str) → Any#

Loads a model from a file.

Parameters:

filepath (str) – The path to the file from which the model will be loaded.

Returns:

An instance of the model.

Return type:

Any

property params: Dict[str, Any]#

Property to retrieve the fitted model parameters. Same as get_params(), but allows attribute-like access.

Returns:

A dictionary containing attack, defense, home advantage, and correlation parameters.

Return type:

dict

Raises:

ValueError – If the model has not been fitted yet.

predict(home_team: str, away_team: str, max_goals: int = 10) → FootballProbabilityGrid[source]#

Predicts the probability of each scoreline for a given home and away team.

Parameters:

  • home_team (str) – The name of the home team

  • away_team (str) – The name of the away team

  • max_goals (int, optional) – The maximum number of goals to consider, by default 10

Returns:

A FootballProbabilityGrid object containing the probability of each scoreline

Return type:

FootballProbabilityGrid

save(filepath: str)#

Saves the model to a file using pickle.

Parameters:

filepath (str) – The path to the file where the model will be saved.