Predict with PHP

1. Install.

composer require tangram/tangram

2. Predict.

First, import the tangram library and load the model file. Then, make an associative array with info for a new patient that matches the CSV, excluding the diagnosis column. Finally, call predict and print out the result.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
<?php namespace tangram\tangram; require_once(dirname(dirname(__FILE__)) . '/vendor/autoload.php'); $model_path = dirname(dirname(__FILE__)) . '/heart_disease.tangram'; $model = Model::from_path($model_path); // Create an example input matching the schema of the CSV file the model was trained on. // Here the data is just hard-coded, but in your application you will probably get this from a database or user input. $input = [ 'age' => 63.0, 'gender' => 'male', 'chest_pain' => 'typical angina', 'resting_blood_pressure' => 145.0, 'cholesterol' => 233.0, 'fasting_blood_sugar_greater_than_120' => 'true', 'resting_ecg_result' => 'probable or definite left ventricular hypertrophy', 'exercise_max_heart_rate' => 150.0, 'exercise_induced_angina' => 'no', 'exercise_st_depression' => 2.3, 'exercise_st_slope' => 'downsloping', 'fluoroscopy_vessels_colored' => '0', 'thallium_stress_test' => 'fixed defect' ]; $output = $model->predict($input); var_dump($output);