Intervals

Confidence and prediction intervals for uncertainty quantification.

Overview

Confidence and Prediction Intervals

Type	Represents	Width	Use
Confidence	Uncertainty in mean curve	Narrow	Where is the true trend?
Prediction	Uncertainty for new points	Wide	Where will new data fall?

Confidence Intervals

Estimate uncertainty in the smoothed curve itself.

const { Loess } = require('fastloess-wasm');

const n = 100;
const x = Float64Array.from({ length: n }, (_, i) => i * 2 * Math.PI / (n - 1));
const y = Float64Array.from(x, (xi, i) => Math.sin(xi) + (((i * 7 + 3) % 17) / 17 - 0.5) * 0.6);

const model = new Loess({ fraction: 0.5, intervals: { confidence : 0.95 } });
const result = model.fit(x, y);

result.y.slice(0, 3).forEach((y, i) => {
    console.log(`x=${result.x[i].toFixed(4)}: y=${y.toFixed(4)} [${result.confidence_lower[i].toFixed(4)}, ${result.confidence_upper[i].toFixed(4)}]`);
});
console.log(`... (${result.y.length - 3} more)`);

x=0.0000: y=0.1118 [-0.0075, 0.2311]
x=0.0635: y=0.1389 [0.0204, 0.2575]
x=0.1269: y=0.1702 [0.0524, 0.2879]
... (97 more)

Prediction Intervals

Estimate where new observations might fall.

const { Loess } = require('fastloess-wasm');

const n = 100;
const x = Float64Array.from({ length: n }, (_, i) => i * 2 * Math.PI / (n - 1));
const y = Float64Array.from(x, (xi, i) => Math.sin(xi) + (((i * 7 + 3) % 17) / 17 - 0.5) * 0.6);

const model = new Loess({ fraction: 0.5, intervals: { prediction : 0.95 } });
const result = model.fit(x, y);
console.log(`Prediction bounds: [${result.prediction_lower[0]}, ${result.prediction_upper[0]}]`);

Prediction bounds: [-0.40481758453074823, 0.6283485157406197]

Both Intervals

Request both types simultaneously:

const { Loess } = require('fastloess-wasm');

const n = 100;
const x = Float64Array.from({ length: n }, (_, i) => i * 2 * Math.PI / (n - 1));
const y = Float64Array.from(x, (xi, i) => Math.sin(xi) + (((i * 7 + 3) % 17) / 17 - 0.5) * 0.6);

const model = new Loess({
    fraction: 0.5,
    intervals: { confidence : 0.95, prediction : 0.95 }
});
const result = model.fit(x, y);
console.log("CI lower[0]:", result.confidence_lower[0].toFixed(4));

CI lower[0]: -0.0075

Confidence Levels

Common levels and their z-values:

Level	z-value	Interpretation
0.90	1.645	90% of intervals contain true value
0.95	1.960	95% of intervals contain true value
0.99	2.576	99% of intervals contain true value

const { Loess } = require('fastloess-wasm');

const n = 100;
const x = Float64Array.from({ length: n }, (_, i) => i * 2 * Math.PI / (n - 1));
const y = Float64Array.from(x, (xi, i) => Math.sin(xi) + (((i * 7 + 3) % 17) / 17 - 0.5) * 0.6);

// 99% confidence interval
const model = new Loess({ intervals: { confidence : 0.99 } });
const result = model.fit(x, y);
console.log("CI lower[0]:", result.confidence_lower[0].toFixed(4));

CI lower[0]: 0.0071

Standard Errors

Access standard errors directly (available when intervals are computed):

const { Loess } = require('fastloess-wasm');

const n = 100;
const x = Float64Array.from({ length: n }, (_, i) => i * 2 * Math.PI / (n - 1));
const y = Float64Array.from(x, (xi, i) => Math.sin(xi) + (((i * 7 + 3) % 17) / 17 - 0.5) * 0.6);

const model = new Loess({ intervals: { confidence : 0.95 } });
const result = model.fit(x, y);

result.standard_errors.slice(0, 5).forEach((se, i) => {
    console.log(`Point ${i}: SE = ${se.toFixed(4)}`);
});
console.log(`... (${result.standard_errors.length - 5} more)`);

Point 0: SE = 0.0618
Point 1: SE = 0.0616
Point 2: SE = 0.0615
Point 3: SE = 0.0614
Point 4: SE = 0.0613
... (95 more)

Availability

Feature	Batch	Streaming	Online
Confidence intervals	✓	✓	✓ (`update_mode: "full"` only)
Prediction intervals	✓	✓	✓ (`update_mode: "full"` only)
Standard errors	✓	✓	✓ (`update_mode: "full"` only)