Simulate from the predictive distribution

Share:

Description

Simulate N- replicates from the predictive distribution for a given time point (tpt) from 1 to n (length of the data).

Usage

1
pred.dist.simul(hyperest, tpt, include.obs = T, N = 1)

Arguments

hyperest

Output from the staircase.hyper.est functions, containing estimates of all hyperparameters

tpt

A specific time point - from 1 to n corresponding to the number of time points from the data set

include.obs

If TRUE, the observed data for time tpt, are also returned

N

Number of replicates

Value

A matrix with N rows; the number of columns depends on whether the observed data are returned

The columns are organized consistent with the observed data (ie. u\times p ungauged blocks, g1\times p, g_2\times p, ...)

Note

This function could be slow if there are missing data at gauged sites correspondind to the selected time point. That is, it is fastest at time points corresponding to Block 1 and slower with higher blocks.