| Title: | Bayesian Ammi Model for Continuous Data |
|---|---|
| Description: | Flexible multi-environment trials analysis via MCMC method for Additive Main Effects and Multiplicative Model (AMMI) for continuous data. Biplot with the averages and regions of confidence can be generated. The chains run in parallel on Linux systems and run serially on Windows. |
| Authors: | Luciano A. Oliveira [aut], Carlos P. Silva [aut], Cristian T.
E. Mendes [aut], Alessandra Q. Silva [aut], Joel J. Nuvunga
[aut], Marcio Balestre [ths], Julio S. S. Bueno-Filho [ths],
Fabio M. Correa [aut, cre]
|
| Maintainer: | Fabio M. Correa <[email protected]> |
| License: | GPL (>= 2) |
| Version: | 1.0-2 |
| Built: | 2024-10-30 06:51:51 UTC |
| Source: | CRAN |
Run the AMMI Bayesian model for continuous data.
ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iterations=3000, jump=2, burn=500, Var.error=0.5, Var.env=0.5, Var.gen=0.5, chains=2)ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iterations=3000, jump=2, burn=500, Var.error=0.5, Var.env=0.5, Var.gen=0.5, chains=2)
Y |
Response variable vector |
Gen |
Genotype effects vector. Must be defined as factor |
Env |
Environmental effects vector. Must be defined as factor |
Rep |
Repetition vector. Must be defined as factor |
iterations |
Total of iterations after burnin and jumo |
jump |
Jump of iterations |
burn |
Initial burn |
Var.error |
Priori for the variance of error. Default is 0.5 |
Var.env |
Priori for the variance of environment. Default is 0.5 |
Var.gen |
Priori for the variance of genotype. Default is 0.5 |
chains |
Number of chains. See details. |
The code is run in parallel for linux SO. If you are using Windows, the execution of the code will be serially.
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) summary(model)library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) summary(model)
Plot the confidence regions for genotype and environment effects
ammiBayes.conf.plot(model, conf=0.95, pars.gen=NULL, pars.env=NULL, gen.labels=NULL, env.labels=NULL, col.env="red", col.gen="green", alpha.env=80, alpha.gen=80, col.text.env="black", col.text.gen="black", border.gen="transparent", border.env="transparent", cex.env=1, cex.gen=1, lty.gen=1, lty.env=1, lwd.gen=1, lwd.env=1, xlab, ylab, col.grid="grey", lty.grid=2, lwd.grid=1, change.signal=FALSE, plot.gen=TRUE, plot.env=TRUE)ammiBayes.conf.plot(model, conf=0.95, pars.gen=NULL, pars.env=NULL, gen.labels=NULL, env.labels=NULL, col.env="red", col.gen="green", alpha.env=80, alpha.gen=80, col.text.env="black", col.text.gen="black", border.gen="transparent", border.env="transparent", cex.env=1, cex.gen=1, lty.gen=1, lty.env=1, lwd.gen=1, lwd.env=1, xlab, ylab, col.grid="grey", lty.grid=2, lwd.grid=1, change.signal=FALSE, plot.gen=TRUE, plot.env=TRUE)
model |
An object of the ammiBayes class |
conf |
Significant level for the confidence region. By default is 0.95. |
pars.gen |
An optional character vector of genotype names. If pars is omitted all genotypes are included. |
pars.env |
An optional character vector of environment names. If pars is omitted all environments are included. |
gen.labels |
Optional vector for the name of the genotypes. |
env.labels |
Optional vector for the name of the environments. |
col.env |
Color for the confidence region of the environment. Default is "red". |
col.gen |
Color for the confidence region of the genotype. Default is "green". |
alpha.env |
Specifies the opacity of the confidence region for the environment. Default is 80. |
alpha.gen |
Specifies the opacity of the confidence region for the genotype. Default is 80. |
col.text.env |
Define the color of environment names. |
col.text.gen |
Define the color of genotype names. |
border.gen |
Define the color for the border of the confidence region of genotype. Default is "transparent". |
border.env |
Define the color for the border of the confidence region of environment. Default is "transparent". |
cex.env |
Scale for the font size of the environment names. Default is 1 |
cex.gen |
Scale for the font size of the genotype names. Default is 1 |
lty.gen |
Line type for the border of confidence region of genotype. Default is 1 |
lty.env |
Line type for the border of confidence region of environment. Default is 1 |
lwd.gen |
Line width for the border of confidence region of genotype. Default is 1 |
lwd.env |
Line width for the border of confidence region of environment. Default is 1 |
xlab |
Label for the x-axis |
ylab |
Label for the y-axis |
col.grid |
Define the color for the grid. Default is "grey" |
lty.grid |
Line type of grid. Default is 2 |
lwd.grid |
Line width of grid. Default is 1 |
change.signal |
Changes the signal of the chain for better visualization of the sample. By default is FALSE |
plot.gen |
Plot effects of genotypes. By default is TRUE |
plot.env |
PLot effects of environment. By default is TRUE |
The confidence regions are defined using the package distfree.cr.
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) ammiBayes.conf.plot(model)library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) ammiBayes.conf.plot(model)
Plot the posterior mean for an ammiBayes object
ammiBayes.gen.plot(x, lwd=1, lty=1, pch=18, method="bars", col.bands=NULL, ylim=NULL, xlab=NULL, ylab=NULL, gen.names=NULL)ammiBayes.gen.plot(x, lwd=1, lty=1, pch=18, method="bars", col.bands=NULL, ylim=NULL, xlab=NULL, ylab=NULL, gen.names=NULL)
x |
An object from gen.effects function. |
lwd |
A line width, a positive number, default is 1. |
lty |
The line type. Default is 1. |
pch |
Either an integer specifying a symbol or a single character to be used as the default in plotting points. |
method |
Defaults to |
col.bands |
Define the color of genotype bands.See |
ylim |
A numeric vector of length 2 giving minimum and maximum for the y-axis. |
xlab |
Label for the x-axis. |
ylab |
Label for the y-axis. |
gen.names |
Define the names of genotypes on the x-axis. By default are the levels of the Genotypes. |
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) genot.effects <- gen.effects(model) ammiBayes.gen.plot(genot.effects)library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) genot.effects <- gen.effects(model) ammiBayes.gen.plot(genot.effects)
Plot the means for the ammiBayes object
ammiBayes.mean.plot(model, pars.gen=NULL, pars.env=NULL, gen.labels=NULL, env.labels=NULL, col.text.gen="darkgreen", col.text.env="red", ylim=NULL, xlim=NULL, cex.env=1, cex.gen=1, xlab, ylab, col.grid="grey", lty.grid=2, lwd.grid=1)ammiBayes.mean.plot(model, pars.gen=NULL, pars.env=NULL, gen.labels=NULL, env.labels=NULL, col.text.gen="darkgreen", col.text.env="red", ylim=NULL, xlim=NULL, cex.env=1, cex.gen=1, xlab, ylab, col.grid="grey", lty.grid=2, lwd.grid=1)
model |
An object of the ammiBayes class |
pars.gen |
An optional character vector of genotype names. If pars is omitted all genotypes are included. |
pars.env |
An optional character vector of environment names. If pars is omitted all environments are included. |
gen.labels |
Optional vector for the name of the genotypes |
env.labels |
Optional vector for the name of the environments |
col.text.gen |
Define the color of genotype names |
col.text.env |
Define the color of environment names |
ylim |
Define the limites applied to the y-axis |
xlim |
Define the limites applied to the x-axis |
cex.env |
Scale for the font size of the environment names. Default is 1 |
cex.gen |
Scale for the font size of the genotype names. Default is 1 |
xlab |
Label for the x-axis |
ylab |
Label for the y-axis |
col.grid |
Define the color for the grid. Default is "grey" |
lty.grid |
Line type of grid |
lwd.grid |
Line width of grid |
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) ammiBayes.mean.plot(model)library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) ammiBayes.mean.plot(model)
Simulated dataset in completely randomized design to illustrate the resources of the ammiBayes package.
data(ammiData)data(ammiData)
amb = Environment (4 environments)
rep = Repetition (9 repetitions)
gen = Genotype (6 genotypes)
prod = Variabel response
Extract the MCMC chain for diagnosis
diagnosis.ammiBayes(x, pars=NULL)diagnosis.ammiBayes(x, pars=NULL)
x |
An object of class ammiBayes |
pars |
It should be set, such as "Genotype", "Rep", "L", "Gen.PC1", "Gen.PC2", "Env.PC1", "Env.PC2", "Comp.var". See details |
The output is compatible for diagnosis with the coda and bayesplot packages.
Examples can be seen on the website: bayesplot
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
# Not run library(ammiBayes) library(bayesplot) library(ggpubr) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=1000, burn=10, jump=2, chains=2) gen.diagnosis <- diagnosis.ammiBayes(model, pars="Genotype") mcmc_trace(gen.diagnosis) mcmc_dens_overlay(gen.diagnosis) mcmc_areas(gen.diagnosis) dens <- bayesplot::mcmc_dens_overlay(gen.diagnosis) trac <- bayesplot::mcmc_trace(gen.diagnosis, facet_args=list(ncol=1)) ggpubr::ggarrange(trac,dens, common.legend=TRUE)# Not run library(ammiBayes) library(bayesplot) library(ggpubr) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=1000, burn=10, jump=2, chains=2) gen.diagnosis <- diagnosis.ammiBayes(model, pars="Genotype") mcmc_trace(gen.diagnosis) mcmc_dens_overlay(gen.diagnosis) mcmc_areas(gen.diagnosis) dens <- bayesplot::mcmc_dens_overlay(gen.diagnosis) trac <- bayesplot::mcmc_trace(gen.diagnosis, facet_args=list(ncol=1)) ggpubr::ggarrange(trac,dens, common.legend=TRUE)
Extract the effects of genotypes and HPD interval
gen.effects(x, prob=0.95)gen.effects(x, prob=0.95)
x |
An object of class ammiBayes |
prob |
Probability for HPD interval. Default is 0.95 |
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) gen.effects(model)library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) gen.effects(model)
Extract predict values and HPD interval
## S3 method for class 'ammiBayes' predict(object, prob=0.95, ...)## S3 method for class 'ammiBayes' predict(object, prob=0.95, ...)
object |
An object of class ammiBayes |
prob |
Probability for HPD interval. Default is 0.95 |
... |
Potential further arguments (required by generic). |
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.
library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) predict(model)library(ammiBayes) data(ammiData) Env <- factor(ammiData$amb) Rep <- factor(ammiData$rep) Gen <- factor(ammiData$gen) Y <- ammiData$prod model <- ammiBayes(Y=Y, Gen=Gen, Env=Env, Rep=Rep, iter=10, burn=1, jump=2, chains=2) predict(model)
Returns (and prints) a summary list for ammiBayes object.
## S3 method for class 'ammiBayes' summary(object, ...)## S3 method for class 'ammiBayes' summary(object, ...)
object |
A given object of the class |
... |
Potential further arguments (required by generic). |
Luciano A. Oliveira
Carlos P. Silva
Cristian T. E. Mendes
Alessandra Q. Silva
Joel J. Nuvunga
Marcio Balestre
Julio S. S. Bueno-Filho
Fabio M. Correa
OLIVEIRA,L.A.; SILVA, C. P.; NUVUNGA, J. J.; SILVA, A. Q.; BALESTRE, M. Credible Intervals for Scores in the AMMI with Random Effects for Genotype. Crop Science, v. 55, p. 465-476, 2015. doi: https://doi.org/10.2135/cropsci2014.05.0369
SILVA, C. P.; OLIVEIRA, L. A.; NUVUNGA, J. J.; PAMPLONA, A. K. A.; BALESTRE, M. A Bayesian Shrinkage Approach for AMMI Models. Plos One, v. 10, p. e0131414, 2015. doi: https://doi.org/10.1371/journal.pone.0131414.