nuget
LmDatumError is an interface for evaluating error, Jacobian matrix and Hessian matrix of a single piece of observed data
LmDatumError
01 Syntax
Microsoft.VisualBasic.Data.Bootstrapping.LevenbergMarquardt.LmDatumError
02 Methods
| Name | Overloads | Summary |
|---|---|---|
| eval | 1 | Evaluates value of the error function for the k-th observed data that corresponds to the parameter vector |
| jacobian | 1 | Evaluates the Jacobian vector of the error function for the k-th observed data that corresponds to the parameter vector |
| hessian | 2 | Evaluates the Hessian matrix of the error function for the k-th observed data that corresponds to the parameter vector |
03 Properties
| Name | Overloads | Summary |
|---|---|---|
| NumData | 1 | Gets the total number of observed data |
04 Members
eval(
Int32, Double())Evaluates value of the error function for the k-th observed data that corresponds to the parameter vector
Parameters
| Name | Type | Description |
|---|---|---|
dataIdx | Int32 | The index of the input data |
optParams | Double() | A vector of real values of parameters in the model @return |
jacobian(
Int32, Double())Evaluates the Jacobian vector of the error function for the k-th observed data that corresponds to the parameter vector
Parameters
| Name | Type | Description |
|---|---|---|
dataIdx | Int32 | The index of the input data |
optParams | Double() | A vector of real values of parameters in the model @return |
hessian(
Int32, Double(), Boolean)Evaluates the Hessian matrix of the error function for the k-th observed data that corresponds to the parameter vector
Parameters
| Name | Type | Description |
|---|---|---|
dataIdx | Int32 | The index of the input data |
optParams | Double() | A vector of real values of parameters in the model |
approxHessianFlg | Boolean | A boolean flag to indicate whether the Hessian matrix can be approximated instead of having to be computed exactly @return |
hessian(
Int32, Double())Evaluates the Hessian matrix of the error function for the k-th observed data that corresponds to the parameter vector. The Hessian matrix is computed exactly
Parameters
| Name | Type | Description |
|---|---|---|
dataIdx | Int32 | - |
optParams | Double() | @return |
NumData
Gets the total number of observed data
Returns
An integer which is the number of observed data