tlars_cpp {tlars}

Exposes the C++ class tlars_cpp to R

Description

Type 'tlars_cpp' in the console to see the constructors, variables, and methods of the class tlars_cpp.

Arguments

X

Real valued predictor matrix.

y

Response vector.

verbose

Logical. If TRUE progress in computations is shown.

intercept

Logical. If TRUE an intercept is included.

standardize

Logical. If TRUE the predictors are standardized and the response is centered.

num_dummies

Number of dummies that are appended to the predictor matrix.

type

Type of used algorithm (currently possible choices: 'lar' or 'lasso').

lars_state

Input list that was extracted from a previous tlars_cpp object using get_all().

T_stop

Number of included dummies after which the random experiments (i.e., forward selection processes) are stopped.

early_stop

Logical. If TRUE, then the forward selection process is stopped after T_stop dummies have been included. Otherwise the entire solution path is computed.

Value

No return value. Exposes the C++ class tlars_cpp to R.

Fields

Constructor:

new - Creates a new object of the class tlars_cpp.

Constructor:

new - Re-creates an object of the class tlars_cpp based on a list of class variables that is obtained via get_all().

Method:

execute_lars_step - Executes LARS steps until a stopping-condition is satisfied.

Method:

get_beta - Returns the estimate of the beta vector.

Method:

get_beta_path - Returns a a matrix with the estimates of the beta vectors at all steps.

Method:

get_num_active - Returns the number of active predictors.

Method:

get_num_active_dummies - Returns the number of dummy variables that have been included.

Method:

get_num_dummies - Returns the number of dummy predictors.

Method:

get_actions - Returns the indices of added/removed variables along the solution path.

Method:

get_df - Returns the degrees of freedom at each step which is given by number of active variables (+1 if intercept is true).

Method:

get_R2 - Returns the R^2 statistic at each step.

Method:

get_RSS - Returns the residual sum of squares at each step.

Method:

get_Cp - Returns the Cp-statistic at each step.

Method:

get_lambda - Returns the lambda-values (penalty parameters) at each step along the solution path.

Method:

get_entry - Returns the first entry/selection steps of the predictors along the solution path.

Method:

get_norm_X - Returns the L2-norm of the predictors.

Method:

get_mean_X - Returns the sample means of the predictors.

Method:

get_mean_y - Returns the sample mean of the response y.

Method:

get_all - Returns all class variables: This list can be used as an input to the constructor to re-create an object of class tlars_cpp.

Examples

data("Gauss_data")
X <- Gauss_data$X
y <- drop(Gauss_data$y)
p <- ncol(X)
n <- nrow(X)
dummies <- matrix(stats::rnorm(n * p), nrow = n, ncol = p)
XD <- cbind(X, dummies)
mod_tlars <- tlars_model(X = XD, y = y, num_dummies = ncol(dummies))
tlars(model = mod_tlars, T_stop = 3, early_stop = TRUE)

mod_tlars$get_beta()

# mod_tlars$get_beta_path()
# mod_tlars$get_num_active()
# mod_tlars$get_num_active_dummies()
# mod_tlars$get_num_dummies()
# mod_tlars$get_actions()
# mod_tlars$get_df()
# mod_tlars$get_R2()
# mod_tlars$get_RSS()
# mod_tlars$get_Cp()
# mod_tlars$get_lambda()
# mod_tlars$get_entry()
# mod_tlars$get_norm_X()
# mod_tlars$get_mean_X()
# mod_tlars$get_mean_y()
# mod_tlars$get_all()