---
title: "OPTICS in JavaScript with @kanaries/ml"
description: "Learn what OPTICS does, when to use it, and how to run OPTICS in JavaScript or TypeScript with @kanaries/ml for browser and Node.js applications."
canonical_url: "https://ml.kanaries.net/docs/apis/clusters/optics"
markdown_url: "https://ml.kanaries.net/docs/apis/clusters/optics.md"
---
# OPTICS in JavaScript

## Algorithm overview

OPTICS captures density-based cluster structure across multiple scales and handles varying-density datasets.

This algorithm is especially useful when:

- DBSCAN-style sensitivity to a single epsilon is too limiting.
- You need ordering information to inspect hierarchical density structure.
- Noise handling is required for reliable segmentation.

## JavaScript implementation

@kanaries/ml makes OPTICS available in JavaScript for density-based clustering workflows where cluster shapes are irregular and noise handling matters. This is useful for web-based data exploration products that need to surface structure without assuming spherical clusters.

Running OPTICS in the JS stack means engineering teams can keep clustering logic close to interactive visualizations, upload flows, or Node.js processing jobs instead of depending on a separate Python layer.

## Interactive OPTICS playground

Tune the reachability extraction radius and minimum neighborhood size. The labels below come from a live `Clusters.OPTICS` fit; click the plot to see whether a new observation joins a density-connected region.

> The HTML version includes a live clustering playground powered by @kanaries/ml. You can change the dataset and algorithm parameters, add observations, and inspect fitted clusters, noise labels, centers, or exemplars. The runnable guide and API reference continue below. [Open the HTML page](https://ml.kanaries.net/docs/apis/clusters/optics).

## Quick start

### OPTICS: Python and JavaScript / TypeScript

The Python example uses scikit-learn; the TypeScript example uses @kanaries/ml in browser or Node.js runtimes.

#### Python (scikit-learn)

```python
from sklearn.cluster import OPTICS

X = [[0, 0], [0.1, 0.1], [5, 5], [5.2, 5.1], [10, 1]]

model = OPTICS(min_samples=2, max_eps=0.6)
labels = model.fit_predict(X)
```

#### JavaScript / TypeScript (@kanaries/ml)

```ts
import { Clusters } from '@kanaries/ml';

const X = [[0, 0], [0.1, 0.1], [5, 5], [5.2, 5.1], [10, 1]];

const model = new Clusters.OPTICS({ min_samples: 2, eps: 0.6 });
const labels = model.fitPredict(X);
```

### Quick JavaScript example

```ts
import { Clusters } from '@kanaries/ml';

const X = [[0, 0], [0.1, 0.1], [5, 5], [5.2, 5.1], [10, 1]];

const model = new Clusters.OPTICS({ min_samples: 2, eps: 0.6 });
const labels = model.fitPredict(X);
console.log(labels);
```

## Detailed API reference

```ts
interface OPTICSOptions {
    min_samples?: number;
    max_eps?: number;
    metric?: Distance.IDistanceType;
    p?: number;
    eps?: number;
}
constructor(options: OPTICSOptions = {})
```

`fitPredict(samplesX: number[][]): number[]` returns cluster labels. Noise points are marked as `-1`.

```ts
const optics = new Clusters.OPTICS({ eps: 0.6, min_samples: 2 });
const labels = optics.fitPredict(X);
```

### Implementation workflow

1. Prepare distance-scaled features and configure neighborhood constraints.
2. Fit and analyze reachability or extracted cluster labels.
3. Tune min-samples and extraction thresholds for your target behavior.

### JavaScript deployment notes

- OPTICS is a better fit than K-Means when your clusters have uneven density or non-convex structure.
- Use it for exploratory clustering and noise detection rather than low-latency request-time scoring.
- Benchmark runtime carefully on larger datasets, especially in browser environments.