AF.Echidna: Summary of added functions for Echidna

In yzhlinscau/AAfun0s: Added Functions for Echidna

Summary of added functions for Echidna

Description

Summary of added functions for Echidna

Usage

get.es0.file(dat.file=NULL,es.file=NULL,
                        path=NULL,message= FALSE,
                        softp=NULL,
                        faS=NULL,pedS=NULL,Rsuffix=FALSE)

echidna(fixed,random,residual,
                   trait,family,weights, 
                   es0.file,softp,
                   delf,foldN,
                   trace,maxit,
                   Fmv,mu.delete,
                   mulT,met,cycle,
                   batch,mulN,mulp,
                   batch.G,batch.R,
                   subF,subV.org,
                   res.no,dat.file,  
                   run.purrr,selfing,
                   predict,vpredict,qualifier,jobqualf)

## S3 method for class 'esR'
update(
  object,
  trait = NULL,
  fixed = NULL,
  random = NULL,
  residual = NULL,
  predict = NULL,
  vpredict = NULL,
  qualifier = NULL,
  jobqualf = NULL,
  trace = NULL,
  maxit = 30,
  selfing = NULL,
  mu.delete = NULL,
  mulN = NULL,
  mulT = NULL,
  met = NULL,
  cycle = NULL,
  softp = NULL,
  batch = NULL,
  batch.G = NULL,
  batch.R = NULL,
  subF = FALSE,
  subV.org = NULL,
  res.no = NULL,
  dat.file = NULL,
  delf = NULL,
  foldN = NULL,
  ...
)

subF(fixed,random,residual,es0.file,
            subV.org, subV.nL,subV.new,mulN,res.no)

b2s(object)

esRT(path,trace=FALSE,mulT=FALSE,met=FALSE,cycle=FALSE)

raneff.acc(object,ran.eff,Var, ped=NULL)

## S3 method for class 'esR'
predict(object)

## S3 method for class 'esR'
coef(object)

## S3 method for class 'esR'
wald(object)

## S3 method for class 'esR'
waldT(object, term = NULL, ncol = NULL)

## S3 method for class 'esR'
IC(object)

## S3 method for class 'esR'
trace(object)

## S3 method for class 'esR'
converge(object)

Arguments

dat.file	data file to generate .es file.
es.file	the .es file to generate .es0 file.
path	the path for data files.
message	show running procedure,FALSE(default).
softp	the path for Echidna software.
fixed	fixed effects, such as, c('Rep'), c('Site', 'Site.Rep') or 'Site Site.Rep', h3~1+Rep, etc.
random	random effects, such as,'Mum','Mum Mum.Rep',~Mum+Mum:Rep, etc.
residual	residual effects, such as,'units','ar1(row).ar1(col)',~ar1(row):ar1(col), etc.
trait	aim trait for analysis, such as, 'h3', 'h3 h4',~h3+h4, etc, NULL(default).
family	such as esr_binomial(), esr_poisson().
weights	A variable used as weights in the fit.
es0.file	the .es0 file.
delf	delete all Echidna result files from the folder of .es0 file, TRUE(default).
foldN	new folder name to store each run's results, only works when delf is 'FALSE'.
trace	show iteration procedure,TRUE(default).
maxit	maximum number of iterations, 30(default).
Fmv	make missing values into fixed terms, FALSE(default).
mu.delete	delete term mu or Trait from model, FALSE(default).
mulT	multi-trait model,FALSE(default).
met	multi-environment trial model,FALSE(default).
cycle	Echidna result from qualifier cycle,FALSE(default).
batch	run batch analysis for more than two trait at one time, FALSE(default).
mulN	trait number for multi-trait analysis at one time, 2(default).
mulp	multi-pin formula to run at one time, NULL(default).
batch.G	run more than two G structures at one time, FALSE(default).
batch.R	run more than two R structures at one time, FALSE(default).
subF	run subF function for MET data sets,FALSE(default).
subV.org	original variable for subF.
res.no	number to show results.
run.purrr	using purrr packages for batch analysis,FALSE(default).
selfing	the probability of selfing for parent, such as 0.1.
predict	prediction for model terms.
vpredict	run vpredict statements with Echidna soft.
qualifier	model qualifiers, such as '!extra 5'.
jobqualf	header line qualifiers, mainly '!view'.
object	Echidna result object in R.

Details

This package would supply some functions for Echidna. Details as following:

Function	Description
get.es0.file	generate .es0 file.
echidna	run mixed models.
wald	output wald results.
Var	output variance components.
summary	output summary results.
IC	output AIC and BIC values.
pin	run pin functions.
predict	output predict results.
plot	output model diagnose results.
coef	output fixed and random effects.
update	update mixed models.
b2s	transform batch esR results to single esR.
model.comp	Model comparison for different mixed models.

Author(s)

Yuanzhen Lin <[e-mail]>

References

Yuanzhen Lin. R & ASReml-R Statistics. China Forestry Publishing House. 2016
Gilmour, A.R. (2020) Echidna Mixed Model Software www.EchidnaMMS.org

Examples

## Not run: 

 library(AFEchidna)

 
 ## generate .es0 file
 get.es0.file(dat.file='fm.csv')
 get.es0.file(es.file='fm.es')
 # file.edit('fm.es0')

res<-echidna(trait='h3',
              fixed='Rep',random='Fam',
              residual=NULL,predict=c('Fam'),
              es0.file="fm.es0")

## method 2                           
# res<-echidna(fixed=h3~1+Rep,random=~Fam,
#              residual=NULL,predict=c('Fam'),
#              es0.file="fm.es0")

 names(res)
 class(res)

 # model diagnose
 plot(res) 

 # wald result
 wald(res)
 waldT(res, term=c('mu','Rep'))
 

 # variance components
 Var(res)

 # summary result
 summary(res)

 # AIC,BIC result
 IC(res)

 # fixed and random effects
 coef(res)$fixed
 coef(res)$random

 # predict results if using predict functions
 mm<-predict(res)
 mm$pred

 # show vc results by using vpredict statements
 pin(res)

 # run pin function to count genetic parameters
 pin11(res,h2~V1/(V1+V2))
 pin(res,mulp=c(h2~V1/(V1+V2),h2f~V1/(V1+V2/4)))

 # model converge stage
 trace(res)
 res$Converge


## End(Not run)

yzhlinscau/AAfun0s documentation built on April 18, 2023, 4:11 p.m.