Bayesian Posterior Plot Generation

class SAES.plots.Pplot.Pplot(data: DataFrame, metrics: DataFrame, metric: str, bayesian_test: str = 'sign')

Bases: object

Class to generate plots for the posterior distribution of a Bayesian statistical test.

Attributes:

data (pd.DataFrame):

A pandas DataFrame containing the performance results of different algorithms across multiple instances.

maximize (bool):

A boolean indicating whether the metric is higher is better.

algorithms (np.array):

An array containing the names of the algorithms under evaluation.

metric (str):

The metric to be used for comparison

logger (Logger):

A logger object to record and display log messages.

Methods:

plot(alg1: str, alg2: str, width: int = 5) -> None:

Plots the posterior distribution of the Bayesian statistical test between two algorithms.

save(alg1: str, alg2: str, output_path: str, file_name: str = None, width: int = 5) -> None:

Saves the posterior distribution of the Bayesian statistical test between two algorithms to a file.

plot_pivot(algorithm: str, width: int = 30) -> None:

Plots the posterior distribution of the Bayesian statistical test between an algorithm and all other algorithms.

save_pivot(algorithm: str, output_path: str, file_name: str = None, width: int = 30) -> None:

Saves the posterior distribution of the Bayesian statistical test between an algorithm and all other algorithms to a file.

save(alg1, alg2, output_path: str, file_name: str = None, width: int = 5, sample_size: int = 2500) → None

Saves the posterior distribution of the Bayesian statistical test between two algorithms to a file.

Args:

alg1 (str):

The name of the first algorithm.

alg2 (str):

The name of the second algorithm.

output_path (str):

The path where the file will be saved.

file_name (str):

The name of the file. Default is None.

width (int):

The width of the figure. Default is 5.

sample_size (int):

Total number of random_search samples generated. Default is 2500.

Returns:

None

Example:

>>> frtom SAES.plots.Pplot import Pplot
>>> import os
>>> 
>>> data = pd.read_csv("data.csv")
>>> metrics = pd.read_csv("metrics.csv")
>>> metric = "HV"
>>> pplot = Pplot(data, metrics, metric)
>>> pplot.save("NSGAII", "OMOPSO", os.getcwd())

save_pivot(algorithm: str, output_path: str, file_name: str = None, width: int = 30, heigth: int = 15, sample_size: int = 2500) → None

Saves the posterior distribution of the Bayesian statistical test between an algorithm and all other algorithms to a file.

Args:

algorithm (str):

The name of the algorithm.

output_path (str):

The path where the file will be saved.

file_name (str):

The name of the file. Default is None.

width (int):

The width of the figure. Default is 30.

heigth (int):

The heigth of the figure. Default is 15.

sample_size (int):

Total number of random_search samples generated. Default is 2500.

Returns:

None

Example:

>>> frtom SAES.plots.Pplot import Pplot
>>> import os
>>> 
>>> data = pd.read_csv("data.csv")
>>> metrics = pd.read_csv("metrics.csv")
>>> metric = "HV"
>>> pplot = Pplot(data, metrics, metric)
>>> pplot.save_pivot("NSGAII", os.getcwd())

show(alg1: str, alg2: str, width: int = 5, sample_size: int = 2500) → None

Plots the posterior distribution of the Bayesian statistical test between two algorithms.

Args:

alg1 (str):

The name of the first algorithm.

alg2 (str):

The name of the second algorithm.

width (int):

The width of the figure. Default is 5.

sample_size (int):

Total number of random_search samples generated. Default is 2500.

Returns:

None

Example:

>>> frtom SAES.plots.Pplot import Pplot
>>> 
>>> data = pd.read_csv("data.csv")
>>> metrics = pd.read_csv("metrics.csv")
>>> metric = "HV"
>>> pplot = Pplot(data, metrics, metric)
>>> pplot.plot("NSGAII", "OMOPSO")

show_pivot(algorithm: str, width: int = 30, heigth: int = 15, sample_size: int = 2500) → None

Plots the posterior distribution of the Bayesian statistical test between an algorithm and all other algorithms.

Args:

algorithm (str):

The name of the algorithm.

width (int):

The width of the figure. Default is 30.

heigth (int):

The heigth of the figure. Default is 15.

sample_size (int):

Total number of random_search samples generated. Default is 2500.

Returns:

None

Example:

>>> frtom SAES.plots.Pplot import Pplot
>>> 
>>> data = pd.read_csv("data.csv")
>>> metrics = pd.read_csv("metrics.csv")
>>> metric = "HV"
>>> pplot = Pplot(data, metrics, metric)
>>> pplot.plot_pivot("NSGAII")