Class | Description |
---|---|
AbstractKernel |
AbstractKernel is the super class of all kernels.
|
BasicKernel |
BasicKernel is a concrete implementation of
DiscreteKernel . |
Binarise |
Binarise kernel: not a convolution kernel.
|
BiWeight |
BiWeight kernel: K(u) = 15/16 * (1 - u^2)^2,
where |u| ≤
1 . |
ContinuousKernel |
ContinousKernel is the super class of all continuous kernels.
|
Cosine |
Cosine kernel: K(u) = Π/4 * cos((Π/2) * u),
where |u| ≤
1 . |
DiscreteKernel |
DiscreteKernel is the super class of all discrete kernels.
|
Epanechnikov |
Epanechnikov kernel: K(u) = 3/4 * (1 - u^2),
where |u| ≤
1 . |
ExponentialMovingAverage |
ExponentialMovingAverage kernels take into account previous values when
determining the current one, but weight those using some exponentially
decreasing weight, or
ExponentialMovingAverage.FACTOR . |
Gaussian |
Gaussian kernel:
K(u) = 1 / (2Π * σ^2) *
e^-((u - μ)^2 / (2 * σ^2)),
where μ =
0 (μ = mu, σ = sigma). |
MovingAverage |
MovingAverage kernels calculate the output value based on all
preceding input values read so far.
|
SimpleMovingAverage |
SimpleMovingAverage kernels calculate the output value based on the previous
and current value(s), using some configurable window
width . |
SimpleMovingMedian |
SimpleMovingMedian kernels calculate the output value based on the previous
and current value(s), using some configurable window
width . |
Triangular |
Triangular kernel: K(u) = 1 - |u|,
where |u| ≤
1 . |
TriCube |
TriCube kernel: K(u) = 70/81 * (1 - u^3)^3,
where |u| ≤
1 . |
TriWeight |
TriWeight kernel: K(u) = 35/32 * (1 - u^2)^3,
where |u| ≤
1 . |
Uniform |
Uniform kernel: K(u) = 1/2,
where |u| ≤
1 . |
WeightedMovingAverage |
WeightedMovingAverage kernels take into account previous values when
determining the current one, but weigh those using configurable weights.
|
Enum | Description |
---|---|
Binarise.TestPolicy |
TestPolicy controls the kind of test the
Kernel performs
on the threshold . |
OutOfBoundsPolicy |
OutOfBoundsPolicy
enum s are used to as default values for
'windowed' operations, when such a window can not, as a whole, be placed over
the input values. |