est {qgen} | R Documentation |
estimating parameters from data sets
Description
estimates variance components for random effects and coefficients
of the fixed effects for a given paraDATA
–object.
Usage
est(paraDATA, file=TRUE, path="~/qgen/")
Arguments
paraDATA |
list with a definded structure; see paraDATA |
file |
Logical flag: Should the paraDATA object be written to
path/"est.rda"? |
path |
indicating the path where the file est.rda should be written |
Details
The parameters of the the
and the
emp
function determine the estimations that are performed and
the exact structure of the output
Value
a paraDATA
–object:
orig |
full |
supl |
full |
para |
full, with phS=NULL |
spec |
depending on the supl -slot of the paraDATA and on the
hist ory of the object itself:
- unbal
- information on the unbalacedness of the
data; different degrees of freedom:
- xxDF
- normal degrees of freedom
- xxDFappREML
- Satterthwaite approximated degrees of
freedom using the REML variance components
- xxDFappANOVA
- Satterthwaite approximated degrees of
freedom using the ANOVA estimators for variance
components
- xxDFappANOVAuw
- Satterthwaite approximated degrees of
freedom using the ANOVA estimators with unweighted sums
of squares for variance components
weights used to calculate the expected mean squares
w1u - for dam variance in calculation of sire mean square
w2u - for sire variance
w3u - for dam variance in calculation of dam mean square
- modelsummary
- additional information:
sBLUP - the best linear unbiased predictors for the sire
effect
modelsumry - the summary(lme4-object)
- secondcontrast
- the same model with a different contrast
matrix to estimate the difference in on character between
environments:
FIX - an array with the means of all fixed effects
modelsumry2 - the summary(lme4-object) for a second
model estimated with contrast for testing whether a character
differs between two environments
secondpartitioning - variance components
additionally (to REML) estimated by:
ANOVA - estimators (normal):
rbS - variance component for factor sire
siS - variance component for factor sire
daS - variance component for factor dam
idS - variance component for residuals
phS NULL
error - 0; the model residuals are included in the idS
ANOVAuw - estimators based on unweighted sums of
squares (for balanced data equal to ANOVA):
rbS - variance component for factor sire
siS - variance component for factor sire
daS - variance component for factor dam
idS - variance component for residuals
phS NULL
error - 0; the model residuals are included in the idS
|
Note
name:est from estimation
References
Burdick, R. K. and Graybill, F. A. (1992)
Confidence intervals on variance components. Marcel Dekker.
Sen, B., Graybill, F. A. and Ting, N. (1992)
Confidence intervals on variance components. Biometrical Journal
3, 259–274.
See Also
paraDATA
Examples
parameters <- est(sim(the(file=FALSE)), file=FALSE)
# takes the default causal parameters of the()
# simulates a data set
# estimates the parameters
[Package
qgen version 0.03-02
Index]