F.cr.model.matrix {mra}

Capture-Recapture model matrix

Description

Returns two model matrices for capture-recapture modeling. Both are in the form of (giant) 2D matrices.

Usage

 
F.cr.model.matrix(capture, survival, nan, ns)

Arguments

capture

Formula for the capture model. Must be a formula object with no response, then ~, followed by the names of 2-D arrays of covariates to fit in the capture model. For example: capture = ~ age + sex, where age and sex are matrices.

survival

Formula for the survival model. Must be a formula object with no response, then ~, followed by the names of 2-D arrays of covariates to fit in the survival model. For example: capture = ~ age + sex, where age and sex are matrices.

nan

Number of individuals in the model. This is necessary for the ivar and tvar functions to work. Normally, nan = number of rows in capture history matrix. No default value.

ns

Number of sampling occasions. Normally, ns = number of columns in the capture history matrix.

Details

This routine is intended to be called internally by model fitting routines of MRA. General users should never have to call this routine.

This routine uses a call to eval with a model frame, and calls the R internal model.matrix to resolve the matrices in the formula. All matrices specified in the models should be in the current scope and accessible to both eval and model.matrix.

This routine calls F.3d.model.matrix twice. F.3d.model.matrix does all the work.

Value

A list containing the following components:

capX

A NAN by IX+(NX*NS) matrix containing covariate values for the capture model. Matrices specified in the model are column appended together. NAN = nrow(x) where x is a 2-D matrix in the model (i.e., number of animals). NS = ncol(x) (i.e., number of capture occasions). NX = number of matrices specified in the model. IX = 1 if an intercept is included, 0 otherwise. The j-th covariate matrix specified in the model can be accessed directly with capX[, IX+1+(NS*(j-1)):(NS*j) ].

surX

A NAN by IY+(NY*NS) matrix containing covariate values for the survival model. Matrices specified in the model are column appended together. NAN = nrow(x) where y is a 2-D matrix in the model (i.e., number of animals). NS = ncol(y) (i.e., number of capture occasions). NY = number of matrices specified in the model. IY = 1 if an intercept is included, 0 otherwise. The j-th covariate matrix specified in the model can be accessed directly with capY[, IY+1+(NS*(j-1)):(NS*j) ].

n.cap.covars

Number of matrices specified in the capture model (NX above).

n.sur.covars

Number of matrices specified in the survival model (NY above).

cap.intercept

TRUE or FALSE depending on whether an intercept was included in the capture model

sur.intercept

TRUE or FALSE depending on whether an intercept was included in the survival model

cap.vars

Vector of names for the NX covariates in the capture model.

sur.vars

Vector of names for the NY covariates in the survival model.

Author(s)

Trent McDonald, WEST-INC, [e-mail]

See Also

F.cjs.estim, model.matrix, eval

Examples

 


# Synthetic example with 10 animals and 5 occasions
nan <- 10
ns <- 5
sex <- matrix( as.numeric(runif( nan ) > 0.5), nrow=nan, ncol=ns )
x <- matrix( runif( nan*ns ) , nrow=nan, ncol=ns )
F.cr.model.matrix( capture= ~ sex + x, survival= ~ -1 + x, nan, ns )