MeanComparison

MeanComparison

The function conducts a test to compare means of two samples. It implies both parametric t-test and non-parametric Kruskal-Wallis H-test.

Parameters:

• x_train_0: pd.Series()
  A pd.Series of a feature of 0 group.
• x_train_1: pd.Series()
  A pd.Series of a feature of 2 group.
• equal_var: bool, optional, default = False
  An indicator of the assumption of equal variance.
• alternative: str, {'two-sided', 'less', 'greater'}, optional, default = 'two-sided'
  Type of alternative hypothesis.

Returns:

• final_data: dict
• Keys: ttest
kruskal
  
• Values: tt: tuple(statistic, pvalue) - results of conducted t-test.
  kruskal: tuple(statistic, pvalue) - results of conducted Kruskal-Wallis H-test.
  

Exceptions:

• TypeError:
  Raised if x_train_0 parameter is not a pd.Series.
  Raised if x_train_1 parameter is not a pd.Series.
  Raised if equal_var parameter is not logical.
  

• ValueError:
  Raised if alternative parameter is not in ('two-sided', 'less', 'greater').
  

Example:

import pandas as pd
from combat.short_list import MeanComparison

# Sample input data
x_train_0 = pd.Series([1, 2, 3, 4, 5])
x_train_1 = pd.Series([6, 7, 8, 9, 10])
equal_var = False
alternative = 'two-sided'

# Perform mean comparison
result = MeanComparison(x_train_0, x_train_1, equal_var, alternative)
print(result)