ridgeCPredict#

Purpose#

Predicts responses using the output from ridgeCFit() and matrix of independent variables.

Format#

predictions = ridgeCPredict(mdl, x_test)#

Returns:

mdl (struct) –

An instance of a ridgeModel structure. An instance named mdl will have the following members:

mdl.alpha_hat	(1 x nlambdas vector) The estimated value for the intercept for each provided lambda.
mdl.beta_hat	(P x nlambdas matrix) The estimated parameter values for each provided lambda.
mdl.mse_train	(nlambdas x 1 vector) The mean squared error for each set of parameters, computed on the training set.
mdl.lambda	(nlambdas x 1 vector) The lambda values used in the estimation.

predictions (Nx1 numeric or string vector) – The predictions.

Parameters:

x_test (NxP matrix) – The independent variables.

Examples#

new;
library gml;

rndseed 23423;

// Create file name with full path
fname = getGAUSSHome("pkgs/gml/examples/breastcancer.csv");

// Load all variables from dataset, except for 'ID'
X = loadd(fname, ". -ID");

// Remove rows with a missing value
X = packr(X);

// Separate dependent and independent variables
y = X[., "class"];
X = delcols(X, "class");

// Split data into 70% training and 30% test set
{ X_train, X_test, y_train, y_test } = trainTestSplit(X, y, 0.7);

// Declare 'mdl' to be an 'ridgeModel' structure
// to hold the trained model
struct ridgeModel mdl;

// Set lambda vector
lambda = seqm(90, 0.4, 5);

// Train the ridge classifier with default settings
mdl = ridgeCFit(y_train, X_train, lambda);

// Make predictions on the test set, from our trained model
y_hat = ridgeCPredict(mdl, X_test);

// Assess model quality
print "";
print "Classification metrics for lambda = "$+ntos(lambda[2]);
call classificationMetrics(y_test, y_hat[.,2]);

print "";
print "Classification metrics for lambda = "$+ntos(lambda[4]);
call classificationMetrics(y_test, y_hat[.,4]);

The code above will print the following output:

===========================================================================
Model:                        Ridge     Target Variable:              class
Number observations:            478     Number features:                  9
===========================================================================

======================================================================
              Lambda   90.0000   36.0000   14.4000    5.7600    2.3040
======================================================================

         clump_thick    0.0026    0.0059    0.0124    0.0225    0.0344
         c_size_unif    0.0027    0.0061    0.0127    0.0219    0.0306
        c_shape_unif    0.0028    0.0063    0.0131    0.0228    0.0323
       marg_adhesion    0.0024    0.0055    0.0113    0.0192    0.0258
     single_epi_size    0.0032    0.0073    0.0149    0.0252    0.0337
         bare_nuclei    0.0023    0.0053    0.0112    0.0208    0.0332
     bland_chromatin    0.0031    0.0071    0.0147    0.0257    0.0369
    normal_nulcleoli    0.0024    0.0055    0.0113    0.0195    0.0275
             mitosis    0.0023    0.0052    0.0102    0.0159    0.0175
              CONST.   -0.3833   -0.4783   -0.6586   -0.9190   -1.1786
======================================================================
        Training MSE     0.818     0.714     0.542     0.349     0.221

Classification metrics for lambda = 36
===================================================
                             Classification metrics
===================================================
       Class   Precision  Recall  F1-score  Support

           0        0.65    1.00      0.79      131
           1        1.00    0.04      0.08       74

   Macro avg        0.82    0.52      0.43      205
Weighted avg        0.78    0.65      0.53      205

    Accuracy                          0.65      205

Classification metrics for lambda = 5.76
===================================================
                             Classification metrics
===================================================
       Class   Precision  Recall  F1-score  Support

           0        0.92    0.98      0.95      131
           1        0.97    0.85      0.91       74

   Macro avg        0.95    0.92      0.93      205
Weighted avg        0.94    0.94      0.94      205

    Accuracy                          0.94      205