statsmodels-js
Version:
basic statistics library.
48 lines (39 loc) • 1.11 kB
JavaScript
import { meanSquaredError, r2Score } from "./metrics";
describe("meanSquaredError", () => {
test("", () => {
const yTrue = [1];
const yPred = [2, 3, 4];
expect(() => {
meanSquaredError(yTrue, yPred);
}).toThrowError(new Error("The size of both arrays should be same."));
});
test("", () => {
const yTrue = [3, -0.5, 2, 7];
const yPred = [2.5, 0.0, 2, 8];
expect(meanSquaredError(yTrue, yPred)).toEqual(0.375);
});
test("", () => {
const yTrue = [3, -0.5, 2, 7];
const yPred = [2.5, 0.0, 2, 8];
expect(meanSquaredError(yTrue, yPred, { squared: false })).toEqual(
0.6123724356957945
);
});
});
describe("r2Score", () => {
test("", () => {
const yTrue = [1, 2];
const yPred = [1, 2];
expect(r2Score(yTrue, yPred)).toEqual(1);
});
test("", () => {
const yTrue = [1, 2];
const yPred = [0.9, 2.1];
expect(r2Score(yTrue, yPred)).toEqual(0.96);
});
test("", () => {
const yTrue = [3, -0.5, 2, 7];
const yPred = [2.5, 0.0, 2, 8];
expect(r2Score(yTrue, yPred)).toEqual(0.9486081370449679);
});
});