SK_EBLUP_LME.f: Evaluate fitted taper curve

In TapeR: Flexible Tree Taper Curves Based on Semiparametric Mixed Models

Evaluate fitted taper curve

Description

This is the actual function to estimate diameters according to the fitted mixed B-splines model.

Usage

SK_EBLUP_LME.f(xm, ym, xp, par.lme, Rfn = list(fn = "sig2"), ...)

Arguments

xm	relative heights for which measurements are available
ym	corresponding diameter measurements in height xm
xp	relative heights for which predictions are required
par.lme	Fitted model object, return of TapeR_FIT_LME.f
Rfn	list with function name to provide estimated or assumed residual variances for the given measurements, optionally parameters for such functions
...	not currently used

Details

This function is the actual working horse for prediction using the fitted taper model. Based on the model par.lme and the measured diameters ym and corresponding (relative) heights xm of a specific tree (there might be just one measurement), the random effect parameters and subsequently diameters are estimated. Depending on the parameter Rfn, the calibrated taper curve is forced through the given diameter ym (Rfn = list(fn="zero")), or calibrated using the complete residual variance-covariance information (Rfn = list(fn="sig2"), the default). Further assumptions are possible, see also resVar and Kublin et al. (2013) p. 987 for more details.

Value

a list holding nine elements:

• b_fix fixed effects parameter of taper model

• b_rnd random effects parameter given tree (posterior mean b_k)

• yp estimated diameter in height xp

• KOV_Mean variance-covariance matrix of expected value

• KOV_Pred variance-covariance matrix of prediction

• CI_Mean mean and limits of confidence interval

• MSE_Mean mean squared error of expected value

• MSE_Pred mean squared error of prediction

• CI_Pred mean and limits of prediction interval

Author(s)

Edgar Kublin

See Also

E_DHx_HmDm_HT.f, E_VOL_AB_HmDm_HT.f, resVar

Examples

data("SK.par.lme")
TapeR:::SK_EBLUP_LME.f(1.3/27, 30, 1.3/27, SK.par.lme)
## using empirical best linear unbiased estimator: estimate != 30
TapeR:::SK_EBLUP_LME.f(1.3/27, 30, 1.3/27, SK.par.lme, Rfn=list(fn="sig2"))$yp
## interpolate / force through given diameter: estimate  == 30
TapeR:::SK_EBLUP_LME.f(1.3/27, 30, 1.3/27, SK.par.lme, Rfn=list(fn="zero"))$yp
TapeR:::SK_EBLUP_LME.f(1.3/27, 30, c(1.3, 5)/27, SK.par.lme)
par.lme <- SK.par.lme
h <- 12 # tree height
xm <- c(1.3, 3) / h # relative measuring height
ym <- c(8, 7.5) # measured diameter
xp <- c(0.5, 1) / h # relative prediction height
TapeR:::SK_EBLUP_LME.f(xm, ym, xp, SK.par.lme)

TapeR documentation built on Aug. 16, 2023, 9:07 a.m.