| vcDelta | R Documentation |
vcDelta method for class "mmes".
Post-analysis procedure to calculate linear combinations of variance components. Its intended use is when the variance components are either simple variances or are variances and covariances in an unstructured matrix. The functions covered are linear combinations of the variance components (for example, phenotypic variance), a ratio of two components (for example, heritabilities) and the correlation based on three components (for example, genetic correlation).
The calculations are based on the estimated variance parameters and their variance matrix as represented by the inverse of the Fisher or Average information matrix. Note that this matrix has zero values for fixed variance parameters including those near the parameter space boundary.
The transform is specified with a formula. On the left side of the formula is a name for the transformation. On the right side of the formula is a transformation specified with shortcut names like 'V1', 'V2', etc. The easiest way to identify these shortcut names is to use 'summary(object)$varcomp'. The rows of this object can referred to with shortcuts 'V1', 'V2', etc. See the example below.
vcDelta(object, transform)
## S3 method for class 'mmes'
vcDelta(object, transform)
## S3 method for class 'mmes.nonlinear'
vcDelta(object, transform)
object |
a model fitted with the mmes function. |
transform |
a formula to calculate the function. |
Nonlinear fits use the observed marginal-likelihood covariance from
post_nonlinear, not linear Fisher/average-information formulas.
Their V1, V2, ... map to natural covariance, residual and
family nuisance entries in coef(object, type="all"), excluding
fixed/curve parameters. The returned parameters attribute records
the mapping. A parameter ratio is not automatically a response-scale
heritability or reliability for a nonlinear model.
The delta method (e.g., Lynch and Walsh 1998, Appendix 1; Ver Hoef 2012) uses
a Taylor series expansion to approximate the moments of a function of
parameters. A first-order delta-method approximation is used for the
standard error of a function of (co)variance parameters.
Partial first derivatives of the function are calculated by algorithmic
differentiation with deriv.
Estimates use the stored native-scale covariance parameter table, and
uncertainty uses its full covariance matrix, including covariance between
different fitted terms. Parameters with zero derivatives do not contribute
uncertainty to the requested transformation.
Though vcDelta can calculate standard errors for non-linear functions
of (co)variance parameters from a fitted mmes model, it is limited
to non-linear functions constructed by mathematical operations such as the
arithmetic operators +, -, *, / and ^,
and single-variable functions such as exp and log. See
deriv for more information.
dd |
the parameter and its standard error. |
Giovanny Covarrubias
Covarrubias-Pazaran G (2016) Genome assisted prediction of quantitative traits using the R package sommer. PLoS ONE 11(6): doi:10.1371/journal.pone.0156744
Lynch, M. and B. Walsh 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Inc., Sunderland, MA, USA.
Ver Hoef, J.M. 2012. Who invented the delta method? The American Statistician 66:124-127. DOI: 10.1080/00031305.2012.687494
vcDelta, mmes
####=========================================####
####=========================================####
#### EXAMPLE 1
#### simple example with univariate models
####=========================================####
####=========================================####
# data(DT_cpdata, package="enhancer")
# DT <- DT_cpdata
# GT <- GT_cpdata
# MP <- MP_cpdata
# #### create the variance-covariance matrix
# A <- A.mat(GT)
# #### look at the data and fit the model
# head(DT)
# mix1 <- mmes(Yield~1,
# random=~vsm(ism(id),Gu=A),
# data=DT)
# summary(mix1)$varcomp
# #### run the vcDelta function
# vcDelta(mix1, h2 ~ V1 / ( V1 + V2 ) )
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.