```
float kspline(string basis, float sample_pos, float value1, float key_pos1, ...)
```

Samples a curve defined by a series of value/position pairs. This is useful for specifying a 1D data ramp.

```
vector kspline(string basis, float sample_pos, vector value1, float key_pos1, ...)
```

```
vector4 kspline(string basis, float sample_pos, vector4 value1, float key_pos1, ...)
```

Samples a curve defined by a series of vector value/position pairs. This is useful for specifying a color ramp.

If you just want linearly spaced keys, or if you need to vary the basis, use spline instead.

`basis`

, `bases`

These are the same interpolations supported by ramp parameters.

`"constant"`

Maintains each key value until the next key, creating a "stair step" curve.

`"linear"`

Connects the key points with a polyline.

For example, if you specified four values:

spline("linear", t, v0, v1, v2, v3)

…the function returns the height of the orange dot at position `sample_pos`.

`"cubic"`

(or `"catmullrom"`

, `"cspline"`

)

Connect the point values with a Catmull-Rom spline.

Note that the first and last values are outside the sample area to provide the slope of the curve at the second point (at the start of the sample range) and the second-to-last point (at the end of the sample range).

For example, if you specified six values:

spline("catrom", t, v0, v1, v2, v3, v4, v5)

…the function returns the height of the orange dot at position `t`.

(This image is for illustration only, it does not show the correct curve for the shown points.)

`"linearsolve"`

(or `"solvelinear"`

)

Maps between a set of non-uniform positions and a set of values. The kspline function does this mapping implicitly.

tk = spline("linearsolve", t, k0, k1, k2, k3, ...); v = spline(basis, tk, v1, v2, v3, ...);

(Technically, `linearsolve`

interprets the values as key values, solves the
intersection of the spline, and returns the intercept point.)

`"monotonecubic"`

`sample_pos`

The position along the curve at which to sample the value.

`value`

, `n``key_pos`

`n`

To define the shape of the curve, you pass a number of value/position pairs specifying the key points through which the curve passes.

You must specify key positions in ascending order or the results will be unpredictable.

Tip

The spline function is a more flexible superset of this function.

This function is the equivalent of:

type kspline(basis, t, v0, k0, v1, k1, v2, k2...) { float tk = spline("linearsolve", t, k0, k1, k2, ...); return spline(basis, tk, v0, v1, v2, ...); }

See also | |

spline |