CRAN Package Check Results for Package bigpca

Last updated on 2018-05-03 10:50:04 CEST.

Flavor Version Tinstall Tcheck Ttotal Status Flags
r-devel-linux-x86_64-debian-clang 1.1 9.48 90.23 99.71 WARN
r-devel-linux-x86_64-debian-gcc 1.1 6.22 77.31 83.53 WARN
r-devel-linux-x86_64-fedora-clang 1.1 129.65 WARN
r-devel-linux-x86_64-fedora-gcc 1.1 122.80 WARN
r-devel-windows-ix86+x86_64 1.1 15.00 83.00 98.00 ERROR
r-patched-linux-x86_64 1.1 6.82 85.75 92.57 WARN
r-patched-solaris-x86 1.1 168.50 WARN
r-release-linux-x86_64 1.1 5.72 85.45 91.17 WARN
r-release-windows-ix86+x86_64 1.1 15.00 83.00 98.00 ERROR
r-release-osx-x86_64 1.1 WARN
r-oldrel-windows-ix86+x86_64 1.1 10.00 135.00 145.00 OK
r-oldrel-osx-x86_64 1.1 OK

Check Details

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/home/hornik/tmp/R.check/r-devel-clang/Work/PKGS/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-devel-linux-x86_64-debian-clang

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/home/hornik/tmp/R.check/r-devel-gcc/Work/PKGS/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-devel-linux-x86_64-debian-gcc

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/data/gannet/ripley/R/packages/tests-clang/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-devel-linux-x86_64-fedora-clang

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/data/gannet/ripley/R/packages/tests-devel/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-devel-linux-x86_64-fedora-gcc

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See 'd:/Rcompile/CRANpkg/local/3.5/bigpca.Rcheck/00install.out' for details.
    Information on the location(s) of code generating the 'Note's can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to 'yes'.
Flavors: r-devel-windows-ix86+x86_64, r-release-windows-ix86+x86_64

Version: 1.1
Check: examples
Result: ERROR
    Running examples in 'bigpca-Ex.R' failed
    The error most likely occurred in:
    
    > ### Name: big.PCA
    > ### Title: PCA/Singular Value Decomposition for big.matrix
    > ### Aliases: big.PCA
    >
    > ### ** Examples
    >
    > # create an example matrix and its transpose
    > min.dim <- 200; nvar <- 500; subset.size <- 50
    > mat <- matrix(rnorm(min.dim*nvar),ncol=min.dim)
    > prv.large(mat)
    Warning in check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    
     col#
     row# 1 2 ..... 200
     1 -0.6265 0.0773 ... 0.0374
     2 0.1836 -0.2969 ... -1.6128
     3 -0.8356 -1.1832 ... 0.7146
     ... ... ... ... ...
     500 -0.8733 -0.6973 ... 0.7118
    > t.mat <- t(mat)
    > # create two alternative covariance matrices
    > MMs <- t.mat %*% mat
    > MsM <- mat %*% t.mat
    > # run singular value decomposition
    > pca <- svd(mat)
    > D <- pca$d # singular values (=sqrt(eigenvalues))
    > V <- pca$v # right singular vector
    > U <- pca$u # left singular vector
    > sig <- mat-mat; diag(sig) <- D;
    > MMs2 <- V %*% (t(sig) %*% sig) %*% t(V)
    > sig <- t.mat-t.mat; diag(sig) <- D;
    > MsM2 <- U %*% (sig %*% t(sig)) %*% t(U)
    > # show that the covariance matrices are equal to the functions of
    > # the left and right singular vectors
    > prv(MMs,MsM); prv(MMs2,MsM2)
    MMs (matrix, 200*200)
    
     col#
     row# 1 2 ..... 200
     1 511.2318 -22.5321 ... -22.6612
     2 -22.5321 558.8836 ... 12.462
     3 -14.5398 -40.9556 ... -32.9189
     ... ... ... ... ...
     200 -22.6612 12.462 ... 568.2967
    
    MsM (matrix, 500*500)
    
     col#
     row# 1 2 ..... 500
     1 198.2148 -1.7057 ... 16.2122
     2 -1.7057 241.7104 ... 0.4053
     3 -14.7584 12.003 ... 4.6714
     ... ... ... ... ...
     500 16.2122 0.4053 ... 191.1251
    
    Warning messages:
    1: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    2: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    3: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    4: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    MMs2 (matrix, 200*200)
    
     col#
     row# 1 2 ..... 200
     1 511.2318 -22.5321 ... -22.6612
     2 -22.5321 558.8836 ... 12.462
     3 -14.5398 -40.9556 ... -32.9189
     ... ... ... ... ...
     200 -22.6612 12.462 ... 568.2967
    
    MsM2 (matrix, 500*500)
    
     col#
     row# 1 2 ..... 500
     1 198.2148 -1.7057 ... 16.2122
     2 -1.7057 241.7104 ... 0.4053
     3 -14.7584 12.003 ... 4.6714
     ... ... ... ... ...
     500 16.2122 0.4053 ... 191.1251
    
    Warning messages:
    1: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    2: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    3: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    4: In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    > pr <- princomp(mat) # PCA using eigendecomposition of cov matrix
    > L <- matrix(rep(0,40000),ncol=200); diag(L) <- pr[[1]]^2 # eigenvalues as diag
    > mat2 <- (pr[[2]]) %*% L %*% solve(pr[[2]]) # = eigenvectors * eigenvalues * inv(eigenvectors)
    > prv.large(cov(mat)); prv.large(mat2) # == COVmat (may be slight tolerance differences)
    
     col#
     row# 1 2 ..... 200
     1 1.024 -0.0441 ... -0.045
     2 -0.0441 1.1179 ... 0.024
     3 -0.0291 -0.0822 ... -0.066
     ... ... ... ... ...
     200 -0.045 0.024 ... 1.1385
    Warning message:
    In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    
     col#
     row# 1 2 ..... 200
     1 1.022 -0.044 ... -0.0449
     2 -0.044 1.1157 ... 0.024
     3 -0.029 -0.082 ... -0.0659
     ... ... ... ... ...
     200 -0.0449 0.024 ... 1.1362
    Warning message:
    In check.bio() :
     bioconductor does not appear to be installed - this function works better if it is
    > ## Now demonstrate the correlation between SVD and PCA ##
    > # the right singular vector is highly correlated with the pca loadings:
    > median(abs(diag(cor(V,pr[["loadings"]]))))
    [1] 0.9834328
    > # the left singular vector is highly correlated with the pca scores (eigenvectors):
    > median(abs(diag(cor(U,pr[["scores"]]))))
    [1] 0.9835682
    > cor(pr$sdev,D) # the singular values are equivalent to the eigenvalues
    [1] 0.9999933
    >
    > ## MAIN EXAMPLES ##
    > orig.dir <- getwd(); setwd(tempdir()); # move to temporary dir
    > if(file.exists("testMyBig.bck")) { unlink(c("testMyBig.bck","testMyBig.dsc")) }
    > bmat <- as.big.matrix(mat,backingfile="testMyBig.bck",
    + descriptorfile="testMyBig.dsc", backingpath = getwd())
    Error in filebacked.big.matrix(nrow = nrow, ncol = ncol, type = type, :
     Backing file already exists! Either remove or specify
     different backing file
    Calls: as.big.matrix ... as.big.matrix -> big.matrix -> filebacked.big.matrix
    Execution halted
Flavors: r-devel-windows-ix86+x86_64, r-release-windows-ix86+x86_64

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/home/hornik/tmp/R.check/r-patched-gcc/Work/PKGS/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-patched-linux-x86_64

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/home/ripley/R/packages/tests32/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-patched-solaris-x86

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/home/hornik/tmp/R.check/r-release-gcc/Work/PKGS/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-release-linux-x86_64

Version: 1.1
Check: whether package can be installed
Result: WARN
    Found the following significant warnings:
     Note: break used in wrong context: no loop is visible
    See ‘/Volumes/SSD-Data/Builds/R-dev-web/QA/Simon/packages/el-capitan-x86_64/results/3.5/bigpca.Rcheck/00install.out’ for details.
    Information on the location(s) of code generating the ‘Note’s can be
    obtained by re-running with environment variable R_KEEP_PKG_SOURCE set
    to ‘yes’.
Flavor: r-release-osx-x86_64