Simulate data to mimic the GPS observations and the DR path.

Share:

Description

Simulate data from our Bayesian melding model with Brownian Bridge and Brownian Motion (See the model description in BMAnimalTrack).

Usage

1
2
dataSim(T, K, s2H, s2D, s2G=0.01, 
  	gind=NULL, betaVec=NULL, dMx=NULL, A=0, B=0, scale=TRUE)

Arguments

T

Number of time points in the animal's path and DR path.

K

Number of GPS observations.

s2H

Variance parameter for Brownian Bridge.

s2D

Variance parameter for the Brownian motion.

s2G

Variance of the measurement error in the GPS observations.

gind

Optional. The time points where the GPS observations are obtained. Default is randomly generating from 1:T.

betaVec

Coefficients in the function h(t). When unspecified, no parametric bias term is considered.

dMx

Design matrix of dimension T. Default the polynomials.

A

Start point of the path. Default 0.

B

End point of the path. Default 0.

scale

Logical (TRUE of FALSE). Whether to standardize the columns of dMx with scale.

Value

A data list with the following elements:

eta

The simulated path of the animal,

Y

The GPS observations,

Ytime

The time points where the GPS observations are available,

X

Dead-Reckoned path

Author(s)

Yang (Seagle) Liu <yang.liu@stat.ubc.ca>

References

Liu, Y., Battaile, B. C., Zidek, J. V., and Trites, A. (2014). Bayesian melding of the Dead-Reckoned path and gps measurements for an accurate and high-resolution path of marine mammals. arXiv preprint arXiv: 1411.6683.

Examples

1
2
3
4
5
6
7
8
set.seed(1)
#Generating data from our 
dlist <- dataSim(T=100, K=10, s2H=1, s2D=0.1, betaVec=c(1))
gpsObs <- dlist$Y
gpsTime <- dlist$Ytime
drPath <- dlist$X
wlist <- as.dataList(drPath, gpsObs, gpsTime, timeUnit=1, s2G=0.01, dUnit=1, betaOrder=1)
##Examples continues in function "as.dataList".