Learn R Programming

spatstat.model (version 3.5-0)

fitted.slrm: Fitted Probabilities for Spatial Logistic Regression

Description

Given a fitted Spatial Logistic Regression model, this function computes the fitted probabilities for each pixel, or the fitted probabilities at each original data point.

Usage

# S3 method for slrm
fitted(object, ..., type="probabilities",
                 dataonly=FALSE, leaveoneout=FALSE)

Value

A pixel image (object of class "im") containing the fitted probability for each pixel, or a numeric vector containing the fitted probability at each data point.

Arguments

object

a fitted spatial logistic regression model. An object of class "slrm".

...

Ignored.

type

Character string (partially) matching one of "probabilities", "intensity" or "link" determining the quantity that should be predicted.

dataonly

Logical. If TRUE, then values will be computed at the points of the data point pattern. If FALSE, then values will be computed at the pixels used to fit the model.

leaveoneout

Logical value specifying whether to perform a leave-one-out calculation. See Details.

Author

Adrian Baddeley Adrian.Baddeley@curtin.edu.au, Rolf Turner rolfturner@posteo.net and Ege Rubak rubak@math.aau.dk.

Details

This is a method for the generic function fitted for spatial logistic regression models (objects of class "slrm", usually obtained from the function slrm).

By default, the fitted presence probabilities in each pixel are computed and returned as a pixel image.

  • If dataonly=FALSE (the default), a value is computed for each pixel in a fine grid.

  • If dataonly=TRUE, a value is computed for each of the points in the original spatial point pattern.

If leaveoneout=FALSE (the default), fitted values are extracted from the fitted model object. If leaveoneout=TRUE, then a leave-one-out calculation is performed:

  • If dataonly=FALSE and leaveoneout=TRUE, the predicted value in each pixel is computed using a leave-one-out calculation. At each pixel j, the model is re-fitted using all the data except the data inside pixel j, and this updated model is then predicted using the data inside pixel j. (‘Leave one pixel out’)

  • If dataonly=TRUE and leaveoneout=TRUE, the predicted value for each data point X[i] is calculated by re-fitting the model to the data with X[i] removed, and then predicting this model at location X[i]. (‘Leave one point out’)

See Also

slrm, fitted

Examples

Run this code
  X <- rpoispp(42)
  fit <- slrm(X ~ x+y)
  plot(fitted(fit))
  fitted(fit, dataonly=TRUE)

Run the code above in your browser using DataLab