crawl (version 1.0-1)

crwPredict: Predict animal locations and velocities using a fitted CTCRW model

Description

The crwPredict function uses a fitted model object from crwMLE to predict animal locations (with estimated uncertainty) at times indicated by the Time column in data and supplimented by times in predTime. If speedEst is set to TRUE, then animal log-speed is also estimated.

Usage

crwPredict(object.crwFit, data, predTime=NULL, speedEst=FALSE, flat=FALSE)

Arguments

object.crwFit

A model object from crwMLE.

data

data used to fit the model (see details in crwMLE.

predTime

vector of additional prediction times.

speedEst

logical. Estimate animal speed or not.

flat

logical. Should the result be returned as a flat data.frame.

Value

• List with the following elements:
• originalData
  A data.frame with is data merged with predTime.
• alpha.hat.y
  A data.frame with predicted state values for each time. First column in latitude location (mu.y), second in velocity (nu.y or theta.y for drift models), and third is drift velocity (gamma.y if specified).
• alpha.hat.x
  longitude state predictions.
• Var.hat.y
  array where Var.hat.y[,,i] is the prediction covariance matrix for alpha.hat.y[,i].
• Var.hat.x
  array or covariance matrices for alpha.hat.x.
• speed
  (If speedEst=TRUE) Gives log speed estimates for each time and standard errors based on delta method. If coordinates are polar, units are meters/unit Time, else, units are those specified by the coordinates.
• If flat is set to TRUE then a data set is returned with the columns of the original data plus the state estimates, standard errors (se) and the speed estimates.

Details

The requirements for data are the same as those for fitting the model in crwMLE.

See Also

See northernFurSeal for additional examples.