VariogramModel
Description
A class for defining variogram model. Variogram models comprise two key components:
Variogram Function:
- this function is fitted to bins of a variogram. Parameters estimated/fixed here are later used by the covariance function.
Covariance Function:
- this function computes the covariance between points based on their distance. It plays a critical role in assigning weights to points with known values during the prediction process.
Note that parameters in variogram and covariance functions must:
- First parameter must be the distance parameter.
- Be ordered consistently, i.e., if parameters for the variogram function are ordered as
distance,sill,range, then parameters for the covariance function must be in the same order.
Methods
set_variogram_func
Set the variogram function.
| Parameter | Accepts | Description |
|---|---|---|
func |
function | The variogram function. |
set_covariance_func
Set the covariance function.
| Parameter | Accepts | Description |
|---|---|---|
func |
function | The covariance function. |
fix_parameter
Fix a parameter of the variogram model to a specific value. Later when the VariogramModel with fixed parameters is
passed to the fit method, this parameter will not be estimated (it will have a constant value). It can also be set
later in the fit method by passing kwargs.
| Parameter | Accepts | Description |
|---|---|---|
param_name |
str | The name of the parameter to fix. |
value |
float | The value to fix the parameter to (argument for the parameter). |
Attributes
variogram_func
The variogram function.
Type: function
covariance_func
The covariance function.
Type: function
args
Tuple containing the names of the arguments of the variogram model.
Type: Tuple[str]
args_numb
Number of arguments of the variogram model.
Type: int
fixed_params
Dictionary containing fixed parameters of the variogram model.
Type: dict