LogisticRegression¶

In [ ]:

LogisticRegression(name: str,
                   penalty: str = 'ENet',
                   tol: float = 1e-4,
                   C: float = 1.0,
                   max_iter: int = 100, 
                   solver: str = 'CGD',
                   l1_ratio: float = 0.5,
                   fit_intercept: bool = True)

Creates a LogisticRegression object using the Vertica LOGISTIC_REG function.

Parameters¶

Name	Type	Optional	Description
name	str	❌	Name of the model to be stored in the database.
penalty	str	✓	Determines the method of regularization. None : No Regularization L1 : L1 Regularization L2 : L2 Regularization ENet : Combination between L1 and L2
tol	float	✓	Determines whether the algorithm has reached the specified accuracy result.
C	float	✓	The regularization parameter value. The value must be zero or non-negative.
max_iter	int	✓	Determines the maximum number of iterations the algorithm performs before achieving the specified accuracy result.
solver	str	✓	The optimizer method to use to train the model. Newton : Newton Method BFGS : Broyden Fletcher Goldfarb Shanno CGD : Coordinate Gradient Descent
l1_ratio	float	✓	ENet mixture parameter that defines how much L1 versus L2 regularization to provide.
fit_intercept	bool	✓	Boolean, specifies whether the model includes an intercept. If False, intercepts are not used when training the model. fit_intercept should not be False when using BFGS optimizer.

Attributes¶

After the object creation, all the parameters become attributes. The model will also create extra attributes when fitting the model:

Name	Type	Description
coef_	tablesample	Coefficients and their mathematical information (pvalue, std, value...)
input_relation	str	Training relation.
X	list	List of the predictors.
y	str	Response column.
test_relation	str	Relation to use to test the model. All model methods are abstractions that simplify the process. The testing relation will be used by the methods to evaluate the model. If empty, the training relation will be used instead. This attribute can be changed at any time.

Methods¶

Name	Description
classification_report / report	Computes a classification report using multiple metrics to evaluate the model (AUC, accuracy, PRC AUC, F1...). In case of multiclass classification, it will consider each category as positive and switch to the next one during the computation.
confusion_matrix	Computes the model confusion matrix.
contour	Draws the model's contour plot.
cutoff_curve	Draws the model Cutoff curve.
deploySQL	Returns the SQL code needed to deploy the model.
drop	Drops the model from the Vertica DB.
features_importance	Computes the model features importance by normalizing the Logistic Regression coefficients.
fit	Trains the model.
lift_chart	Draws the model Lift Chart.
plot	Draws the Logistic Regression if the number of predictors is equal to 1 or 2.
get_attr	Returns the model attribute.
get_params	Returns the model Parameters.
prc_curve	Draws the model precision-recall curve.
predict	Predicts using the input relation.
predict_proba	Returns the model's probabilities using the input relation.
roc_curve	Draws the model ROC curve.
score	Computes the model score.
set_params	Sets the parameters of the model.
to_memmodel	Converts a specified Vertica model to a memModel model.
to_python	Returns the Python code needed to deploy the model without using built-in Vertica functions.
to_sql	Returns the SQL code needed to deploy the model without using Vertica built-in functions.

Example¶

In [4]:

from verticapy.learn.linear_model import LogisticRegression
model = LogisticRegression(name = "public.LR_titanic",
                           tol = 1e-4, 
                           C = 1.0, 
                           max_iter = 100, 
                           solver = 'CGD',
                           l1_ratio = 0.5)
display(model)

<LogisticRegression>