Estimates the probability density of the input variable.
Namespace:
Extreme.Statistics
Assembly:
Extreme.Numerics (in Extreme.Numerics.dll) Version: 8.1.1
public static ContinuousDistribution EstimateDistribution(
Vector<double> input,
Kernel kernel,
double bandwidth = NaN,
KernelDensityBandwidthEstimator bandwidthEstimator = KernelDensityBandwidthEstimator.NormalReference,
double bandwidthAdjustment = 1,
int dataPoints = -1,
double padding = 3
)
Public Shared Function EstimateDistribution (
input As Vector(Of Double),
kernel As Kernel,
Optional bandwidth As Double = NaN,
Optional bandwidthEstimator As KernelDensityBandwidthEstimator = KernelDensityBandwidthEstimator.NormalReference,
Optional bandwidthAdjustment As Double = 1,
Optional dataPoints As Integer = -1,
Optional padding As Double = 3
) As ContinuousDistribution
public:
static ContinuousDistribution^ EstimateDistribution(
Vector<double>^ input,
Kernel^ kernel,
double bandwidth = NaN,
KernelDensityBandwidthEstimator bandwidthEstimator = KernelDensityBandwidthEstimator::NormalReference,
double bandwidthAdjustment = 1,
int dataPoints = -1,
double padding = 3
)
static member EstimateDistribution :
input : Vector<float> *
kernel : Kernel *
?bandwidth : float *
?bandwidthEstimator : KernelDensityBandwidthEstimator *
?bandwidthAdjustment : float *
?dataPoints : int *
?padding : float
(* Defaults:
let _bandwidth = defaultArg bandwidth NaN
let _bandwidthEstimator = defaultArg bandwidthEstimator KernelDensityBandwidthEstimator.NormalReference
let _bandwidthAdjustment = defaultArg bandwidthAdjustment 1
let _dataPoints = defaultArg dataPoints -1
let _padding = defaultArg padding 3
*)
-> ContinuousDistribution
Parameters
- input
- Type: Extreme.MathematicsVectorDouble
- kernel
- Type: Extreme.StatisticsKernel
The kernel to use for the estimation. - bandwidth (Optional)
- Type: SystemDouble
Optional. The bandwidth. - bandwidthEstimator (Optional)
- Type: Extreme.StatisticsKernelDensityBandwidthEstimator
Optional. The method used to compute the bandwidth. - bandwidthAdjustment (Optional)
- Type: SystemDouble
Optional. A factor by which to adjust the bandwidth. - dataPoints (Optional)
- Type: SystemInt32
Optional. The number of data points to use in the approximation. - padding (Optional)
- Type: SystemDouble
Optional. The space before the smallest element in input
and after the largest element in input where the probability density
should be approximated in units of the bandwidth. The default value is 3.
Return Value
Type:
ContinuousDistributionThe estimated distribution.
When bandwidth is supplied, it is used as the bandwidth,
and the value of bandwidthEstimator is ignored.
Otherwise, the method specified by bandwidthEstimator
is used to estimate the bandwidth. The default is to use the normal
reference bandwidth, which minimizes the integrated square error
under the assumption that the data is normal.
The bandwidth is multiplied by the bandwidthAdjustment,
with a default value of 1. This makes it possible to specify a bandwidth
as a fraction of the estimated bandwidth.
The probability density function (PDF) of the distribution is approximated using
dataPoints points. If dataPoints is less than
or equal to zero, then the larger of the length of input and 50 is used.
Reference