CRAN Package Check Results for Package NMF

Last updated on 2019-07-03 01:47:31 CEST.

Check Details

Version: 0.21.0
Flags: --no-vignettes
Check: examples
Result: ERROR
    Running examples in ‘NMF-Ex.R’ failed
    The error most likely occurred in:
    
    > ### Name: NMFfitX-class
    > ### Title: Virtual Class to Handle Results from Multiple Runs of NMF
    > ### Algorithms
    > ### Aliases: NMFfitX-class
    >
    > ### ** Examples
    >
    > ## Don't show:
    > # roxygen generated flag
    > options(R_CHECK_RUNNING_EXAMPLES_=TRUE)
    > ## End(Don't show)
    >
    > # generate a synthetic dataset with known classes
    > n <- 20; counts <- c(5, 2, 3);
    > V <- syntheticNMF(n, counts)
    >
    > # perform multiple runs of one algorithm (default is to keep only best fit)
    > res <- nmf(V, 3, nrun=3)
    # NOTE - CRAN check detected: limiting maximum number of cores [2/16]
    Error in big.matrix(nrow = nrow(x), ncol = ncol(x), type = type, init = NULL, :
     Error: memory could not be allocated for instance of type big.matrix
    Calls: nmf ... gVariable -> <Anonymous> -> <Anonymous> -> big.matrix
    Timing stopped at: 0.033 0.039 0.172
    Execution halted
Flavor: r-patched-solaris-x86

Version: 0.21.0
Check: package dependencies
Result: NOTE
    Packages suggested but not available for checking: 'doMPI', 'Biobase'
Flavor: r-release-osx-x86_64

Version: 0.21.0
Check: R code for possible problems
Result: NOTE
    .foreach_regfun: no visible global function definition for
     ‘getS3method’
    .wrapResult: no visible global function definition for ‘exprs’
    hash_function: no visible global function definition for
     ‘capture.output’
    nmfInfo: no visible global function definition for ‘packageDescription’
    nmfReport : nmfRun: no visible global function definition for ‘str’
    print.foreach_backend: no visible global function definition for ‘str’
    setupLibPaths: no visible global function definition for
     ‘capture.output’
    setupTempDirectory: no visible global function definition for
     ‘file_test’
    staticVar: no visible global function definition for ‘capture.output’
    str_args: no visible global function definition for ‘capture.output’
    txtProgressBar : up1: no visible global function definition for
     ‘flush.console’
    txtProgressBar : up2: no visible global function definition for
     ‘flush.console’
    txtProgressBar : up3: no visible global function definition for
     ‘flush.console’
    txtProgressBar : kill: no visible global function definition for
     ‘flush.console’
    nmf,matrix-numeric-NMFStrategy: no visible global function definition
     for ‘capture.output’
    nmf,matrix-numeric-NMFStrategy : run.all: no visible global function
     definition for ‘setTxtProgressBar’
    nmf,matrix-numeric-NMFStrategy : run.all: no visible global function
     definition for ‘file_test’
    nmfModel,formula-ANY : merge_pdata: no visible global function
     definition for ‘pData’
    nmfModel,formula-ANY: no visible global function definition for ‘exprs’
    seed,ANY-list-NMFSeed: no visible global function definition for
     ‘capture.output’
    show,NMFfit: no visible global function definition for ‘capture.output’
    Undefined global functions or variables:
     capture.output exprs file_test flush.console getS3method pData
     packageDescription setTxtProgressBar str
    Consider adding
     importFrom("utils", "capture.output", "file_test", "flush.console",
     "getS3method", "packageDescription", "setTxtProgressBar",
     "str")
    to your NAMESPACE file.
Flavor: r-release-osx-x86_64

Version: 0.21.0
Check: Rd cross-references
Result: NOTE
    Package unavailable to check Rd xrefs: ‘Biobase’
Flavor: r-release-osx-x86_64

Version: 0.21.0
Check: data for non-ASCII characters
Result: WARN
     Error loading dataset 'esGolub':
     Error in .requirePackage(package) :
     unable to find required package 'Biobase'
    
     The dataset(s) may use package(s) not declared in the DESCRIPTION file.
Flavor: r-release-osx-x86_64

Version: 0.21.0
Check: examples
Result: ERROR
    Running examples in ‘NMF-Ex.R’ failed
    The error most likely occurred in:
    
    > ### Name: nmfModel
    > ### Title: Factory Methods NMF Models
    > ### Aliases: nmfModel nmfModel,data.frame,data.frame-method
    > ### nmfModel,formula,ANY-method nmfModel,matrix,ANY-method
    > ### nmfModel,matrix,matrix-method nmfModel-methods
    > ### nmfModel,missing,ANY-method nmfModel,missing,missing-method
    > ### nmfModel,NULL,ANY-method nmfModel,numeric,matrix-method
    > ### nmfModel,numeric,missing-method nmfModel,numeric,numeric-method
    > ### nmfModels
    > ### Keywords: methods
    >
    > ### ** Examples
    >
    > ## Don't show:
    > # roxygen generated flag
    > options(R_CHECK_RUNNING_EXAMPLES_=TRUE)
    > ## End(Don't show)
    >
    > #----------
    > # nmfModel,numeric,numeric-method
    > #----------
    > # data
    > n <- 20; r <- 3; p <- 10
    > V <- rmatrix(n, p) # some target matrix
    >
    > # create a r-ranked NMF model with a given target dimensions n x p as a 2-length vector
    > nmfModel(r, c(n,p)) # directly
    <Object of class:NMFstd>
    features: 20
    basis/rank: 3
    samples: 10
    > nmfModel(r, dim(V)) # or from an existing matrix <=> nmfModel(r, V)
    <Object of class:NMFstd>
    features: 20
    basis/rank: 3
    samples: 10
    > # or alternatively passing each dimension separately
    > nmfModel(r, n, p)
    <Object of class:NMFstd>
    features: 20
    basis/rank: 3
    samples: 10
    >
    > # trying to create a NMF object based on incompatible matrices generates an error
    > w <- rmatrix(n, r)
    > h <- rmatrix(r+1, p)
    > try( new('NMFstd', W=w, H=h) )
    Error in validObject(.Object) :
     invalid class “NMFstd” object: Dimensions of W and H are not compatible [ncol(W)= 3 != nrow(H)= 4 ]
    > try( nmfModel(w, h) )
    Error in .local(rank, target, ...) :
     nmfModel - Invalid number of columns in the basis matrix [3]: it should match the number of rows in the mixture coefficient matrix [4]
    > try( nmfModel(r+1, W=w, H=h) )
    Error in .local(rank, target, ...) :
     nmfModel - Objective rank [4] is greater than the number of columns in W [3]
    > # The factory method can be force the model to match some target dimensions
    > # but warnings are thrown
    > nmfModel(r, W=w, H=h)
    Warning in .local(rank, target, ...) :
     nmfModel - Objective rank [3] is lower than the number of rows in H [4]: only the first 3 rows of H will be used
    <Object of class:NMFstd>
    features: 20
    basis/rank: 3
    samples: 10
    > nmfModel(r, n-1, W=w, H=h)
    Warning in .local(rank, target, ...) :
     nmfModel - Number of rows in target is lower than the number of rows in W [20]: only the first 19 rows of W will be used
    Warning in .local(rank, target, ...) :
     nmfModel - Objective rank [3] is lower than the number of rows in H [4]: only the first 3 rows of H will be used
    <Object of class:NMFstd>
    features: 19
    basis/rank: 3
    samples: 10
    >
    > #----------
    > # nmfModel,numeric,missing-method
    > #----------
    > ## Empty model of given rank
    > nmfModel(3)
    <Object of class:NMFstd>
    features: 0
    basis/rank: 3
    samples: 0
    >
    > #----------
    > # nmfModel,missing,ANY-method
    > #----------
    > nmfModel(target=10) #square
    <Object of class:NMFstd>
    features: 10
    basis/rank: 0
    samples: 10
    > nmfModel(target=c(10, 5))
    <Object of class:NMFstd>
    features: 10
    basis/rank: 0
    samples: 5
    >
    > #----------
    > # nmfModel,missing,missing-method
    > #----------
    > # Build an empty NMF model
    > nmfModel()
    <Object of class:NMFstd>
    features: 0
    basis/rank: 0
    samples: 0
    >
    > # create a NMF object based on one random matrix: the missing matrix is deduced
    > # Note this only works when using factory method NMF
    > n <- 50; r <- 3;
    > w <- rmatrix(n, r)
    > nmfModel(W=w)
    <Object of class:NMFstd>
    features: 50
    basis/rank: 3
    samples: 0
    >
    > # create a NMF object based on random (compatible) matrices
    > p <- 20
    > h <- rmatrix(r, p)
    > nmfModel(H=h)
    <Object of class:NMFstd>
    features: 0
    basis/rank: 3
    samples: 20
    >
    > # specifies two compatible matrices
    > nmfModel(W=w, H=h)
    <Object of class:NMFstd>
    features: 50
    basis/rank: 3
    samples: 20
    > # error if not compatible
    > try( nmfModel(W=w, H=h[-1,]) )
    Error in .local(rank, target, ...) :
     nmfModel - Invalid number of columns in the basis matrix [3]: it should match the number of rows in the mixture coefficient matrix [2]
    >
    > #----------
    > # nmfModel,numeric,matrix-method
    > #----------
    > # create a r-ranked NMF model compatible with a given target matrix
    > obj <- nmfModel(r, V)
    > all(is.na(basis(obj)))
    [1] TRUE
    >
    > #----------
    > # nmfModel,matrix,matrix-method
    > #----------
    > ## From two existing factors
    >
    > # allows a convenient call without argument names
    > w <- rmatrix(n, 3); h <- rmatrix(3, p)
    > nmfModel(w, h)
    <Object of class:NMFstd>
    features: 50
    basis/rank: 3
    samples: 20
    >
    > # Specify the type of NMF model (e.g. 'NMFns' for non-smooth NMF)
    > mod <- nmfModel(w, h, model='NMFns')
    > mod
    <Object of class:NMFns>
    features: 50
    basis/rank: 3
    samples: 20
    theta: 0.5
    >
    > # One can use such an NMF model as a seed when fitting a target matrix with nmf()
    > V <- rmatrix(mod)
    > res <- nmf(V, mod)
    > nmf.equal(res, nmf(V, mod))
    [1] TRUE
    >
    > # NB: when called only with such a seed, the rank and the NMF algorithm
    > # are selected based on the input NMF model.
    > # e.g. here rank was 3 and the algorithm "nsNMF" is used, because it is the default
    > # algorithm to fit "NMFns" models (See ?nmf).
    >
    > #----------
    > # nmfModel,matrix,ANY-method
    > #----------
    > ## swapped arguments `rank` and `target`
    > V <- rmatrix(20, 10)
    > nmfModel(V) # equivalent to nmfModel(target=V)
    <Object of class:NMFstd>
    features: 20
    basis/rank: 0
    samples: 10
    > nmfModel(V, 3) # equivalent to nmfModel(3, V)
    <Object of class:NMFstd>
    features: 20
    basis/rank: 3
    samples: 10
    >
    > #----------
    > # nmfModel,formula,ANY-method
    > #----------
    > # empty 3-rank model
    > nmfModel(~ 3)
    <Object of class:NMFstd>
    features: 0
    basis/rank: 3
    samples: 0
    >
    > # 3-rank model that fits a given data matrix
    > x <- rmatrix(20,10)
    > nmfModel(x ~ 3)
    <Object of class:NMFstd>
    features: 20
    basis/rank: 3
    samples: 10
    >
    > # add fixed coefficient term defined by a factor
    > gr <- gl(2, 5)
    > nmfModel(x ~ 3 + gr)
    <Object of class:NMFstd>
    features: 20
    basis/rank: 5
    samples: 10
    fixed coef [2]:
     gr = <1, 2>
    >
    > # add fixed coefficient term defined by a numeric covariate
    > nmfModel(x ~ 3 + gr + b, data=list(b=runif(10)))
    <Object of class:NMFstd>
    features: 20
    basis/rank: 6
    samples: 10
    fixed coef [3]:
     gr = <1, 2>
     b = 0.0101301828399301, 0.21454192395322, ..., 0.767450851621106
    >
    > # 3-rank model that fits a given ExpressionSet (with fixed coef terms)
    > e <- ExpressionSet(x)
    Error in ExpressionSet(x) : could not find function "ExpressionSet"
    Execution halted
Flavor: r-release-osx-x86_64

Version: 0.21.0
Check: re-building of vignette outputs
Result: WARN
    Error(s) in re-building vignettes:
     ...
    --- re-building ‘NMF-vignette.Rnw’ using knitr
    Quitting from lines 385-398 (NMF-vignette.Rnw)
    Error: processing vignette 'NMF-vignette.Rnw' failed with diagnostics:
    unable to find required package 'Biobase'
    --- failed re-building ‘NMF-vignette.Rnw’
    
    --- re-building ‘heatmaps.Rnw’ using knitr
    Converted 2 of 2 package citations to BibTeX
    Writing 4 Bibtex entries ... OK
    Results written to file 'Rpackages.bib'
    --- finished re-building ‘heatmaps.Rnw’
    
    SUMMARY: processing the following file failed:
     ‘NMF-vignette.Rnw’
    
    Error: Vignette re-building failed.
    Execution halted
Flavor: r-release-osx-x86_64

Version: 0.21.0
Check: package dependencies
Result: NOTE
    Package suggested but not available for checking: ‘doMPI’
Flavor: r-oldrel-osx-x86_64