Predict with JavaScript
1. Install.
dependencies: {
"@modelfoxdotdev/modelfox": "*",
}
2. Predict.
First, import the modelfox 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
32
33
34
const fs = require("fs");
const path = require("path");
const modelfox = require("@modelfoxdotdev/modelfox");
// Get the path to the .modelfox file.
const modelPath = path.join(__dirname, "heart_disease.modelfox");
// Load the model from the path.
const modelData = fs.readFileSync(modelPath);
const model = new modelfox.Model(modelData.buffer);
// 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.
const input = {
age: 63,
gender: "male",
chest_pain: "typical angina",
resting_blood_pressure: 145,
cholesterol: 233,
fasting_blood_sugar_greater_than_120: "true",
resting_ecg_result: "probable or definite left ventricular hypertrophy",
exercise_max_heart_rate: 150,
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!
const output = model.predict(input);
// Print the output.
console.log("Output:", output);