fit_and_score: Fit and Score

In UBC-MDS/punisheR: Feature Selection Using AIC and BIC

Description

Fits a linear regression model to the data and scores the relative quality of the model using r-squared, aic, or bic.

Usage

fit_and_score(S, feature, algorithm, X_train, y_train, X_val, y_val, criterion)

Arguments

S	A vector representing the list of selected features in 'forward()' and 'backward()'
feature	Feature to add or drop, expressed as an integer.
algorithm	Direction of feature selection. One of: 'forward', 'backward'.
X_train	Training data. Represented as a 2D matrix of (observations, features).
y_train	Target class for training data. Represented as a 1D vector of target classes for X_train.
X_val	Validation data. Represented as a 2D matrix of (observations, features).
y_val	Target class for validation data. Represented as a 1D vector of target classes for X_val.
criterion	Model selection criterion to measure relative model quality. Can be one of: 'aic': use Akaike Information Criterion 'bic': use Akaike Information Criterion 'r-squared': use coefficient of determination

Value

Score of the model (as a float).

UBC-MDS/punisheR documentation built on May 25, 2019, 1:36 p.m.