Simpleline simplifies N dimensional lines using the Ramer–Douglas–Peucker algorithm.

It is a port of https://github.com/mourner/simplify-js by Vladimir Agafonkin, extended to support N dimensions and custom metrics (distance functions).

To use simpleline, you provide a slice of Point objects.

Point is an interface defined by:

type Point interface {
	Vector() []float64
	Scale(float64) Point
	Subtract(Point) Point
	Zero() Point
}

A three dimensional point, Point3d, is provided. You can use use it in 2D if you want.

You simplify your list of points by calling

points := []Point{...}

results, err := simpleline.RDP(points, 5, simpleline.Euclidean, true)
if err != nil {
	// handle error
}

simpleline.RDP is called like:

simpleline.RDP(points, epsilon, metric function, simplify first?)

The Euclidean metric is provided. You can implement your own function that satisfies the Metric interface.

type Metric func(Point, Point) float64

You could get better performance. For example, this implementation uses vector math and calculates the actual distance rather than its square.

Be careful with floats near MaxFloat64.