Predict with Ruby

1. Install.

gem install tangram

2. Predict.

First, import the tangram library and load the model file. Then, make an object 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
31
require 'tangram' # Get the path to the .tangram file. model_path = File.join(File.dirname(__FILE__), 'heart_disease.tangram') # Load the model from the path. model = Tangram::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', } # Make the prediction! output = model.predict(input) # Print the output. puts('Output:', output)