IsModelValid

IsModelValid

The function calculates whether the model meets prespecified requirements.

Parameters:

• model: LogitModel
  An object of LogitModel class.
• coef_expectation: pd.DataFrame
  A pandas DataFrame object containing variable names and their sign expectations.
• p_value: float, default = 0.05
  Variable significance level.
• check_sample: str, {'test', 'train'}, default = 'test'
  The sample to perform the test.
• metric: str, {'gini', 'auc'}, default = 'gini'
  A metric used to measure the model accuracy.
• gini_cutoff: float, default = 0.4
  A cutoff value for Gini.
• auc_cutoff: float, default = 0.7
  A cutoff value for AUC.

Exceptions:

• TypeError: Raised if the model parameter is not of type LogitModel.
  Raised if the coef_expectation parameter is not a pandas DataFrame.
  

• ValueError: Raised if the check_sample parameter is not in ['train', 'test'].
  Raised if the metric parameter is not in ['gini', 'auc'].
  Raised if the p_value parameter is not a float from 0 to 0.5.
  Raised if the gini_cutoff parameter is not a float from 0 to 1.
  Raised if the auc_cutoff parameter is not a float from 0 to 1.
  

Example:

from combat.models import LogitModel
from combat.combat import IsModelValid
import pandas as pd

# Sample input data
model = LogitModel()
coef_expectation = pd.DataFrame({'variable': ['var1', 'var2'], 'sign_expectation': [1, -1]})
p_value = 0.05
check_sample = 'test'
metric = 'gini'
gini_cutoff = 0.4
auc_cutoff = 0.7

# Check if the model is valid
validity = IsModelValid(model, coef_expectation, p_value, check_sample, metric, gini_cutoff, auc_cutoff)
print("Is the model valid?", validity)