MLSplineFit

Performs a spline based line fit to a set of data.

Given a set of training parameters and observations (x and y values) along with a parameter controlling the smoothness of the required output, the function returns a set of values that make up a curve that fits to the parameters and observations.

Arguments

Number Type Compulsory Default Description

1

real array

Yes

Training parameters (x values)

2

real array

Yes

Training observations (y values)

3

real array

Yes

Smoothness parameter

4

real array

Yes

Result parameters (x values)

Argument 1

The parameters for the training data. This would normally be the values on the x-axis of a graph. The values must be ordered from lowest to highest value.

Argument 2

The observations for the training data. This would normally be the values on the y-axis of a graph.

Argument 3

Parameter that controls how smooth the fit to the data will be. Value must be 0-positive, where the smoothness of the fit increases as the parameter increases.

At the extremes, a value of 0 produces a result made up of straight lines between each training point in order, whilst a value tending towards infinity produces a single straight line through the whole of the data.

Argument 4

The parameters to fit the resulting curve to.

Returns

Return type: real array

Vector the same length as parameter 4 (Result parameters (x values)) , with fitted values for each parameter in order.