#!/usr/bin/env perl
## regression8.pl
## This script demoknstrates how can carry out bulk prediction for all the
## test samples in a csv file.
## IMPORTANT: The topmost record in the CSV file for bulk testing must
## state the names of the fields.
## For the example shown below, there is only one predictor variable.
## This is reflected in the test csv records in the bulk testing file
##
## bulk_testing_data.csv
##
## The output results are deposited in the file
##
## bulk_testing_data_output.csv
use strict;
use warnings;
use Algorithm::RegressionTree;
my $training_datafile = "gendata4.csv";
my $rt = Algorithm::RegressionTree->new(
training_datafile => $training_datafile,
dependent_variable_column => 2,
predictor_columns => [1],
mse_threshold => 0.01,
max_depth_desired => 1,
csv_cleanup_needed => 1,
);
$rt->get_training_data_for_regression();
my $root_node = $rt->construct_regression_tree();
print "\n\nThe Regression Tree:\n";
$root_node->display_regression_tree(" ");
print "\n\n";
$rt->bulk_predictions_for_data_in_a_csv_file($root_node, "bulk_testing_data.csv", [1]);