Class for the trained Model object — DistributionModel-class • ibis.iSDM

All trained Models inherit the options here plus any additional ones defined by the engine and inference.

Note

Could be further pretified and commands outsourced.

Public fields

id: A character id for any trained model
name: A description of the model as character.
model: A list containing all input datasets and parameters to the model.
settings: A Settings object with information on inference.
fits: A list containing the prediction and fitted model.
.internals: A list containing previous fitted models.

Methods

Public methods

DistributionModel$new()
DistributionModel$get_name()
DistributionModel$print()
DistributionModel$show()
DistributionModel$plot()
DistributionModel$plot_threshold()
DistributionModel$show_duration()
DistributionModel$summary()
DistributionModel$effects()
DistributionModel$get_equation()
DistributionModel$get_data()
DistributionModel$get_model()
DistributionModel$set_data()
DistributionModel$get_thresholdvalue()
DistributionModel$get_thresholdtype()
DistributionModel$show_rasters()
DistributionModel$get_projection()
DistributionModel$get_resolution()
DistributionModel$rm_threshold()
DistributionModel$calc_suitabilityindex()
DistributionModel$get_centroid()
DistributionModel$has_limits()
DistributionModel$has_latent()
DistributionModel$has_offset()
DistributionModel$mask()
DistributionModel$save()
DistributionModel$clone()

`DistributionModel$new()`

Initializes the object and creates an empty list

Usage

DistributionModel$new(name)

Arguments

name: A description of the model as character.

Returns

NULL

`DistributionModel$get_name()`

Return the name of the model

Usage

DistributionModel$get_name()

Returns

A character with the model name used.

`DistributionModel$print()`

Print the names and summarizes the model within

Usage

DistributionModel$print()

Returns

A message on screen

`DistributionModel$show()`

Show the name of the Model.

Usage

DistributionModel$show()

Returns

A character of the run name.

`DistributionModel$plot()`

Plots the prediction if found.

Usage

DistributionModel$plot(what = "mean")

Arguments

what: character with the specific layer to be plotted.

Returns

A graphical representation of the prediction

`DistributionModel$plot_threshold()`

Plots the thresholded prediction if found.

Usage

DistributionModel$plot_threshold(what = 1)

Arguments

what: character or numeric for the layer to be plotted.

Returns

A graphical representation of the thresholded prediction if found.

`DistributionModel$show_duration()`

Show model run time if settings exist

Usage

DistributionModel$show_duration()

Returns

A numeric estimate of the duration it took to fit the models.

`DistributionModel$summary()`

Get effects or importance tables from model

Usage

DistributionModel$summary(obj = "fit_best")

Arguments

obj: A character of which object to return.

Returns

A data.frame summarizing the model, usually its coefficient.

`DistributionModel$effects()`

Generic plotting function for effect plots

Usage

DistributionModel$effects(x = "fit_best", what = "fixed", ...)

Arguments

x: A character for the object in question.
what: A character for the type of coefficients.
...: Any other options.

Returns

A graphical representation of the coefficents.

`DistributionModel$get_equation()`

Get equation

Usage

DistributionModel$get_equation()

Returns

A formula of the inferred model.

`DistributionModel$get_data()`

Get specific fit from this Model

Usage

DistributionModel$get_data(x = "prediction")

Arguments

x: A character stating what should be returned.

Returns

A SpatRaster object with the prediction.

`DistributionModel$get_model()`

Small internal helper function to directly get the model object

Usage

DistributionModel$get_model()

Returns

A fitted model if existing.

`DistributionModel$set_data()`

Set new fit for this Model.

Usage

DistributionModel$set_data(x, value)

Arguments

x: The name of the new fit.
value: The SpatRaster layer (or model) to be inserted.

Returns

This object.

`DistributionModel$get_thresholdvalue()`

Get the threshold value if calculated

Usage

DistributionModel$get_thresholdvalue()

Returns

A numeric threshold value.

`DistributionModel$get_thresholdtype()`

Get threshold type and format if calculated.

Usage

DistributionModel$get_thresholdtype()

Returns

A vector with a character method and numeric threshold value.

`DistributionModel$show_rasters()`

List all rasters in object

Usage

DistributionModel$show_rasters()

Returns

A vector with logical flags for the various objects.

`DistributionModel$get_projection()`

Get projection of the background.

Usage

DistributionModel$get_projection()

Returns

A geographic projection

`DistributionModel$get_resolution()`

Get the resolution of the projection

Usage

DistributionModel$get_resolution()

Returns

numeric estimates of the distribution.

`DistributionModel$rm_threshold()`

Remove calculated thresholds

Usage

DistributionModel$rm_threshold()

Returns

Invisible

`DistributionModel$calc_suitabilityindex()`

Calculate a suitability index for a given projection

Usage

DistributionModel$calc_suitabilityindex(method = "normalize")

Arguments

method: The method used for normalization.

Details

Methods can either be normalized by the minimum and maximum. Or the relative total using the sumof values.

Returns

Returns a SpatRaster.

`DistributionModel$get_centroid()`

Get centroids of prediction layers

Usage

DistributionModel$get_centroid(patch = FALSE, layer = "mean")

Arguments

patch: A logical if centroid should be calculated weighted by values.
layer: character of the layer to use.

Returns

Returns a sf object.

`DistributionModel$has_limits()`

Logical indication if the prediction was limited.

Usage

DistributionModel$has_limits()

Returns

A logical flag.

`DistributionModel$has_latent()`

Logical indication if the prediction has added latent factors.

Usage

DistributionModel$has_latent()

Returns

A logical flag.

`DistributionModel$has_offset()`

Has a offset been used?

Usage

DistributionModel$has_offset()

Returns

A logical flag.

`DistributionModel$mask()`

Convenience function to mask all input datasets.

Usage

DistributionModel$mask(mask, inverse = FALSE, ...)

Arguments

mask: A SpatRaster or sf object.
inverse: A logical flag if the inverse should be masked instead.
...: Any other parameters passed on to mask

Returns

Invisible

`DistributionModel$save()`

Save the prediction as output.

Usage

DistributionModel$save(fname, type = "gtif", dt = "FLT4S")

Arguments

fname: An output filename as character.
type: A format as character. Matched against a list of supported formats.
dt: The datatype used, such as float64

Returns

Saved spatial prediction on drive.

`DistributionModel$clone()`

The objects of this class are cloneable with this method.

Usage

DistributionModel$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.