Last updated on 2020-08-14 16:48:00 CEST.
| Flavor | Version | Tinstall | Tcheck | Ttotal | Status | Flags |
|---|---|---|---|---|---|---|
| r-devel-linux-x86_64-debian-clang | 1.0-1 | 3.61 | 33.16 | 36.77 | ERROR | |
| r-devel-linux-x86_64-debian-gcc | 1.0-1 | 3.32 | 26.08 | 29.40 | ERROR | |
| r-devel-linux-x86_64-fedora-clang | 1.0-1 | 53.51 | ERROR | |||
| r-devel-linux-x86_64-fedora-gcc | 1.0-1 | 43.52 | ERROR | |||
| r-devel-windows-ix86+x86_64 | 1.0-1 | 12.00 | 65.00 | 77.00 | ERROR | |
| r-patched-linux-x86_64 | 1.0-1 | 4.29 | 32.90 | 37.19 | ERROR | |
| r-patched-solaris-x86 | 1.0-1 | 71.10 | ERROR | |||
| r-release-linux-x86_64 | 1.0-1 | 4.34 | 33.26 | 37.60 | ERROR | |
| r-release-macos-x86_64 | 1.0-1 | NOTE | ||||
| r-release-windows-ix86+x86_64 | 1.0-1 | 11.00 | 46.00 | 57.00 | ERROR | |
| r-oldrel-macos-x86_64 | 1.0-1 | NOTE | ||||
| r-oldrel-windows-ix86+x86_64 | 1.0-1 | 7.00 | 40.00 | 47.00 | ERROR |
Version: 1.0-1
Check: R code for possible problems
Result: NOTE
bayes.regress: no visible binding for global variable 'var'
bayesregressB1S1: no visible global function definition for 'rgamma'
bayesregressB1S2: no visible global function definition for 'rgamma'
bayesregressB2S1: no visible global function definition for 'rgamma'
bayesregressB2S2: no visible global function definition for 'rgamma'
bayesregressB3S1: no visible global function definition for 'rWishart'
bayesregressB3S1: no visible global function definition for 'rgamma'
bayesregressB3S2: no visible global function definition for 'rWishart'
bayesregressB3S2: no visible global function definition for 'rgamma'
read.regress.data.ff: no visible global function definition for
'chunk.ffdf'
Undefined global functions or variables:
chunk.ffdf rWishart rgamma var
Consider adding
importFrom("stats", "rWishart", "rgamma", "var")
to your NAMESPACE file.
Flavors: r-devel-linux-x86_64-debian-clang, r-devel-linux-x86_64-debian-gcc, r-devel-linux-x86_64-fedora-clang, r-devel-linux-x86_64-fedora-gcc, r-devel-windows-ix86+x86_64, r-patched-linux-x86_64, r-patched-solaris-x86, r-release-linux-x86_64, r-release-windows-ix86+x86_64, r-oldrel-windows-ix86+x86_64
Version: 1.0-1
Check: examples
Result: ERROR
Running examples in 'BayesSummaryStatLM-Ex.R' failed
The error most likely occurred in:
> base::assign(".ptime", proc.time(), pos = "CheckExEnv")
> ### Name: bayes.regress
> ### Title: MCMC posterior sampling of Bayesian linear regression model
> ### parameters using only summary statistics
> ### Aliases: bayes.regress
> ### Keywords: combine consensus subposterior posterior parallel
>
> ### ** Examples
>
> ##################################################
> ## Simulate data
> ##################################################
>
> set.seed(284698)
>
> num.samp <- 100 # number of data values to simulate
>
> # The first value of the beta vector is the y-intercept:
> beta <- c(-0.33, 0.78, -0.29, 0.47, -1.25)
>
> # Calculate the number of predictor variables:
> num.pred <- length(beta)-1
>
> rho <- 0.0 # correlation between predictors
> mean.vec <- rep(0,num.pred)
> sigma.mat <- matrix(rho,num.pred,num.pred) + diag(1-rho,num.pred)
> sigmasq.sim <- 0.05
>
> # Simulate predictor variables:
> x.pre <- rmvn(num.samp, mu=mean.vec, sigma=sigma.mat)
>
> # Add leading column of 1's to x.pre for y-intercept:
> x <- cbind(rep(1,num.samp),x.pre)
>
> epsilon <- rnorm(num.samp, mean=0, sd=sqrt(sigmasq.sim))
>
> y <- as.numeric( x %*% as.matrix(beta) + epsilon)
>
> ## Compute summary statistics (alternatively, the
> # "read.regress.data.ff() function (in this package) can be
> # used to calculate summary statistics; see example below).
>
> xtx <- t(x)%*%x
> xty <- t(x)%*%y
> yty <- t(y)%*%y
>
> data.values<-list(xtx=xtx, xty=xty, yty=yty,
+ numsamp.data = num.samp,
+ xtx.inv = chol2inv(chol(xtx)))
>
> ##########################################################
> ## Bayesian linear regression analysis
> ##########################################################
>
> Tsamp.out <- 100 # number of MCMC samples to produce
>
> ## Choose priors for beta and sigma-squared. Here,
> # beta: Uniform prior; sigma-squared: Inverse Gamma prior.
>
> beta.prior <- list( type = "flat")
> sigmasq.prior <- list(type = "inverse.gamma", inverse.gamma.a = 1.0,
+ inverse.gamma.b = 1.0, sigmasq.init = 1.0 )
>
> set.seed(284698)
>
> # Run the "bayes.regress" function using the data simulated above.
>
> MCMC.out <- bayes.regress(data.values,
+ beta.prior,
+ sigmasq.prior = sigmasq.prior,
+ Tsamp.out = Tsamp.out)
>
> # Next, print the posterior means of the unknown model parameters.
> # Alternatively, the "coda" package can be used for analysis.
>
> print(c(colMeans(MCMC.out$beta), mean(MCMC.out$sigmasq)))
[1] -0.31947013 0.83635812 -0.24389683 0.46429934 -1.22235687 0.07309549
>
> # Check that output is close to simulated values (although num.samp and
> # Tsamp.out are small here); note that the output includes both beta and
> # sigmasq:
> # c(-0.33, 0.78, -0.29, 0.47, -1.25, 0.05)
>
> ## Run all 6 combinations of priors for 3 "beta.prior" choices and
> # 2 "sigmasq.prior" choices:
>
> beta.priors <- list(
+ list( type = "flat"),
+
+ list( type = "mvnorm.known",
+ mean.mu = rep(0.0, (num.pred+1)),
+ prec.Cinv = diag(1.0, (num.pred+1))),
+
+ list( type = "mvnorm.unknown",
+ mu.hyper.mean.eta = rep(0.0,(num.pred+1)),
+ mu.hyper.prec.Dinv = diag(1.0, (num.pred+1)),
+ Cinv.hyper.df.lambda = (num.pred+1),
+ Cinv.hyper.invscale.Vinv = diag(1.0, (num.pred+1)),
+ mu.init = rep(1.0, (num.pred+1)),
+ Cinv.init = diag(1.0,(num.pred+1)))
+ )
>
> sigmasq.priors <- list(
+ list(type = "inverse.gamma",
+ inverse.gamma.a = 1.0,
+ inverse.gamma.b = 1.0,
+ sigmasq.init = 0.1 ),
+ list( type="sigmasq.inverse", sigmasq.init = 0.1)
+ )
>
> for (beta.prior in beta.priors)
+ {
+ for(sigmasq.prior in sigmasq.priors)
+ {
+ set.seed(284698)
+ MCMC.out <- bayes.regress(data.values,
+ beta.prior,
+ sigmasq.prior = sigmasq.prior,
+ Tsamp.out = Tsamp.out)
+ print(c(colMeans(MCMC.out$beta), mean(MCMC.out$sigmasq)))
+ }
+ }
[1] -0.31983032 0.83600490 -0.24455739 0.46498638 -1.22276428 0.07281071
[1] -0.32019753 0.83621751 -0.24526931 0.46502564 -1.22328984 0.05331027
[1] -0.31935446 0.83537445 -0.24438076 0.46477280 -1.22204028 0.07281941
[1] -0.31984715 0.83575331 -0.24513920 0.46486854 -1.22275688 0.05331623
[1] -0.31703522 0.83822588 -0.24989106 0.46958822 -1.22709261 0.07244923
[1] -0.31786752 0.83817854 -0.24982952 0.46895645 -1.22705289 0.05304441
>
> # Check that output is close to simulated values (although num.samp and
> # Tsamp.out are small here); note that the output includes both beta and
> # sigmasq:
> # c(-0.33, 0.78, -0.29, 0.47, -1.25, 0.05):
>
>
> #######################################################################
> ## Read the data from a file, calculate the summary statistics and run
> ## the Bayesian linear regression analysis
> #######################################################################
>
> Tsamp.out <- 100
>
> ## Assume non-zero y-intercept data.
>
> # Read the files and compute summary statistics using the "read.regress.data.ff()"
> # function (in this package).
>
>
> filename <- system.file('data/regressiondata.nz.all.csv.gz', package='BayesSummaryStatLM')
> data.values <- read.regress.data.ff(filename)
Error in chunk.ffdf(data.ffdf) : could not find function "chunk.ffdf"
Calls: read.regress.data.ff
Execution halted
Flavors: r-devel-linux-x86_64-debian-clang, r-devel-linux-x86_64-debian-gcc, r-patched-linux-x86_64, r-release-linux-x86_64
Version: 1.0-1
Check: examples
Result: ERROR
Running examples in ‘BayesSummaryStatLM-Ex.R’ failed
The error most likely occurred in:
> ### Name: bayes.regress
> ### Title: MCMC posterior sampling of Bayesian linear regression model
> ### parameters using only summary statistics
> ### Aliases: bayes.regress
> ### Keywords: combine consensus subposterior posterior parallel
>
> ### ** Examples
>
> ##################################################
> ## Simulate data
> ##################################################
>
> set.seed(284698)
>
> num.samp <- 100 # number of data values to simulate
>
> # The first value of the beta vector is the y-intercept:
> beta <- c(-0.33, 0.78, -0.29, 0.47, -1.25)
>
> # Calculate the number of predictor variables:
> num.pred <- length(beta)-1
>
> rho <- 0.0 # correlation between predictors
> mean.vec <- rep(0,num.pred)
> sigma.mat <- matrix(rho,num.pred,num.pred) + diag(1-rho,num.pred)
> sigmasq.sim <- 0.05
>
> # Simulate predictor variables:
> x.pre <- rmvn(num.samp, mu=mean.vec, sigma=sigma.mat)
>
> # Add leading column of 1's to x.pre for y-intercept:
> x <- cbind(rep(1,num.samp),x.pre)
>
> epsilon <- rnorm(num.samp, mean=0, sd=sqrt(sigmasq.sim))
>
> y <- as.numeric( x %*% as.matrix(beta) + epsilon)
>
> ## Compute summary statistics (alternatively, the
> # "read.regress.data.ff() function (in this package) can be
> # used to calculate summary statistics; see example below).
>
> xtx <- t(x)%*%x
> xty <- t(x)%*%y
> yty <- t(y)%*%y
>
> data.values<-list(xtx=xtx, xty=xty, yty=yty,
+ numsamp.data = num.samp,
+ xtx.inv = chol2inv(chol(xtx)))
>
> ##########################################################
> ## Bayesian linear regression analysis
> ##########################################################
>
> Tsamp.out <- 100 # number of MCMC samples to produce
>
> ## Choose priors for beta and sigma-squared. Here,
> # beta: Uniform prior; sigma-squared: Inverse Gamma prior.
>
> beta.prior <- list( type = "flat")
> sigmasq.prior <- list(type = "inverse.gamma", inverse.gamma.a = 1.0,
+ inverse.gamma.b = 1.0, sigmasq.init = 1.0 )
>
> set.seed(284698)
>
> # Run the "bayes.regress" function using the data simulated above.
>
> MCMC.out <- bayes.regress(data.values,
+ beta.prior,
+ sigmasq.prior = sigmasq.prior,
+ Tsamp.out = Tsamp.out)
>
> # Next, print the posterior means of the unknown model parameters.
> # Alternatively, the "coda" package can be used for analysis.
>
> print(c(colMeans(MCMC.out$beta), mean(MCMC.out$sigmasq)))
[1] -0.31947013 0.83635812 -0.24389683 0.46429934 -1.22235687 0.07309549
>
> # Check that output is close to simulated values (although num.samp and
> # Tsamp.out are small here); note that the output includes both beta and
> # sigmasq:
> # c(-0.33, 0.78, -0.29, 0.47, -1.25, 0.05)
>
> ## Run all 6 combinations of priors for 3 "beta.prior" choices and
> # 2 "sigmasq.prior" choices:
>
> beta.priors <- list(
+ list( type = "flat"),
+
+ list( type = "mvnorm.known",
+ mean.mu = rep(0.0, (num.pred+1)),
+ prec.Cinv = diag(1.0, (num.pred+1))),
+
+ list( type = "mvnorm.unknown",
+ mu.hyper.mean.eta = rep(0.0,(num.pred+1)),
+ mu.hyper.prec.Dinv = diag(1.0, (num.pred+1)),
+ Cinv.hyper.df.lambda = (num.pred+1),
+ Cinv.hyper.invscale.Vinv = diag(1.0, (num.pred+1)),
+ mu.init = rep(1.0, (num.pred+1)),
+ Cinv.init = diag(1.0,(num.pred+1)))
+ )
>
> sigmasq.priors <- list(
+ list(type = "inverse.gamma",
+ inverse.gamma.a = 1.0,
+ inverse.gamma.b = 1.0,
+ sigmasq.init = 0.1 ),
+ list( type="sigmasq.inverse", sigmasq.init = 0.1)
+ )
>
> for (beta.prior in beta.priors)
+ {
+ for(sigmasq.prior in sigmasq.priors)
+ {
+ set.seed(284698)
+ MCMC.out <- bayes.regress(data.values,
+ beta.prior,
+ sigmasq.prior = sigmasq.prior,
+ Tsamp.out = Tsamp.out)
+ print(c(colMeans(MCMC.out$beta), mean(MCMC.out$sigmasq)))
+ }
+ }
[1] -0.31983032 0.83600490 -0.24455739 0.46498638 -1.22276428 0.07281071
[1] -0.32019753 0.83621751 -0.24526931 0.46502564 -1.22328984 0.05331027
[1] -0.31935446 0.83537445 -0.24438076 0.46477280 -1.22204028 0.07281941
[1] -0.31984715 0.83575331 -0.24513920 0.46486854 -1.22275688 0.05331623
[1] -0.31703522 0.83822588 -0.24989106 0.46958822 -1.22709261 0.07244923
[1] -0.31786752 0.83817854 -0.24982952 0.46895645 -1.22705289 0.05304441
>
> # Check that output is close to simulated values (although num.samp and
> # Tsamp.out are small here); note that the output includes both beta and
> # sigmasq:
> # c(-0.33, 0.78, -0.29, 0.47, -1.25, 0.05):
>
>
> #######################################################################
> ## Read the data from a file, calculate the summary statistics and run
> ## the Bayesian linear regression analysis
> #######################################################################
>
> Tsamp.out <- 100
>
> ## Assume non-zero y-intercept data.
>
> # Read the files and compute summary statistics using the "read.regress.data.ff()"
> # function (in this package).
>
>
> filename <- system.file('data/regressiondata.nz.all.csv.gz', package='BayesSummaryStatLM')
> data.values <- read.regress.data.ff(filename)
Error in chunk.ffdf(data.ffdf) : could not find function "chunk.ffdf"
Calls: read.regress.data.ff
Execution halted
Flavors: r-devel-linux-x86_64-fedora-clang, r-devel-linux-x86_64-fedora-gcc, r-devel-windows-ix86+x86_64, r-patched-solaris-x86, r-release-windows-ix86+x86_64, r-oldrel-windows-ix86+x86_64
Version: 1.0-1
Check: R code for possible problems
Result: NOTE
bayes.regress: no visible binding for global variable ‘var’
bayesregressB1S1: no visible global function definition for ‘rgamma’
bayesregressB1S2: no visible global function definition for ‘rgamma’
bayesregressB2S1: no visible global function definition for ‘rgamma’
bayesregressB2S2: no visible global function definition for ‘rgamma’
bayesregressB3S1: no visible global function definition for ‘rWishart’
bayesregressB3S1: no visible global function definition for ‘rgamma’
bayesregressB3S2: no visible global function definition for ‘rWishart’
bayesregressB3S2: no visible global function definition for ‘rgamma’
Undefined global functions or variables:
rWishart rgamma var
Consider adding
importFrom("stats", "rWishart", "rgamma", "var")
to your NAMESPACE file.
Flavors: r-release-macos-x86_64, r-oldrel-macos-x86_64