CRAN Package Check Results for Package CompareCausalNetworks

Last updated on 2015-09-06 23:47:19.

Flavor Version Tinstall Tcheck Ttotal Status Flags
r-devel-linux-x86_64-debian-clang 0.1.1 2.45 42.94 45.39 NOTE
r-devel-linux-x86_64-debian-gcc 0.1.1 2.38 43.56 45.94 NOTE
r-devel-linux-x86_64-fedora-clang 0.1.1 90.26 NOTE
r-devel-linux-x86_64-fedora-gcc 0.1.1 85.98 NOTE
r-devel-osx-x86_64-clang 0.1.1 75.68 OK
r-devel-windows-ix86+x86_64 0.1.1 6.00 75.00 81.00 NOTE
r-patched-linux-x86_64 0.1.1 2.51 44.98 47.49 OK
r-patched-solaris-sparc 0.1.1 582.60 ERROR
r-patched-solaris-x86 0.1.1 136.10 OK
r-release-linux-x86_64 0.1.1 2.54 45.09 47.63 OK
r-release-osx-x86_64-mavericks 0.1.1 OK
r-release-windows-ix86+x86_64 0.1.1 9.00 66.00 75.00 OK
r-oldrel-windows-ix86+x86_64 0.1.1 12.00 79.00 91.00 OK

Check Details

Version: 0.1.1
Check: R code for possible problems
Result: NOTE
    bivariateCAM: no visible global function definition for ‘var’
    drawE: no visible global function definition for ‘runif’
    getParentsStable: no visible global function definition for ‘as’
    getParentsStable: no visible global function definition for ‘rnorm’
    indtestHsic: no visible global function definition for ‘dist’
    indtestHsic: no visible global function definition for ‘median’
    indtestHsic: no visible global function definition for ‘qgamma’
    indtestHsic: no visible global function definition for ‘pgamma’
    runCAM: no visible global function definition for ‘as’
    runGES: no visible global function definition for ‘packageVersion’
    runGES: no visible global function definition for ‘new’
    runGES: no visible global function definition for ‘as’
    runGIES: no visible global function definition for ‘new’
    runGIES: no visible global function definition for ‘as’
    runHiddenICP: no visible global function definition for ‘coef’
    runICP: no visible global function definition for ‘coef’
    runPC: no visible global function definition for ‘cor’
    runPC: no visible global function definition for ‘as’
    runRFCI: no visible global function definition for ‘cor’
    runRFCI: no visible global function definition for ‘as’
    runRegression: no visible global function definition for ‘coef’
    train_gam: no visible global function definition for ‘formula’
    Undefined global functions or variables:
     as coef cor dist formula median new packageVersion pgamma qgamma
     rnorm runif var
    Consider adding
     importFrom("methods", "as", "new")
     importFrom("stats", "coef", "cor", "dist", "formula", "median",
     "pgamma", "qgamma", "rnorm", "runif", "var")
     importFrom("utils", "packageVersion")
    to your NAMESPACE (and ensure that your DESCRIPTION Imports field
    contains 'methods').
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

Version: 0.1.1
Check: examples
Result: ERROR
    Running examples in ‘CompareCausalNetworks-Ex.R’ failed
    The error most likely occurred in:
    
    > ### Name: getParents
    > ### Title: Estimate the connectivity matrix of a causal graph
    > ### Aliases: getParents
    > ### Keywords: Causality, Graph estimations
    >
    > ### ** Examples
    >
    > ## load the backShift package for data generation and plotting functionality
    > if(!requireNamespace("backShift", quietly = TRUE))
    + stop("The package 'backShift' is needed for the examples to
    + work. Please install it.", call. = FALSE)
    >
    > require(backShift)
    Loading required package: backShift
    >
    > # Simulate data with connectivity matrix A with assumptions
    > # 1) hidden variables present
    > # 2) precise location of interventions is assumed unknown
    > # 3) different environments can be distinguished
    >
    > ## simulate data
    > myseed <- 1
    >
    > # sample size n
    > n <- 10000
    >
    > # p=3 predictor variables and connectivity matrix A
    > p <- 3
    > labels <- c("1", "2", "3")
    > A <- diag(p)*0
    > A[1,2] <- 0.8
    > A[2,3] <- 0.8
    > A[3,1] <- -0.4
    >
    > # divide data in 10 different environments
    > G <- 10
    >
    > # simulate
    > simResult <- simulateInterventions(n, p, A, G, intervMultiplier = 3,
    + noiseMult = 1, nonGauss = TRUE, hiddenVars = TRUE,
    + knownInterventions = FALSE, fracVarInt = NULL, simulateObs = TRUE,
    + seed = myseed)
    > X <- simResult$X
    > environment <- simResult$environment
    >
    > ## apply all methods given in vector 'methods'
    > ## (using all data pooled for pc/LINGAM/rfci/ges -- can be changed with option
    > ## 'onlyObservationalData=TRUE')
    >
    > methods <- c("backShift", "LINGAM") #c("pc", "rfci", "ges")
    >
    > # select whether you want to run stability selection
    > stability <- FALSE
    >
    > # arrange graphical output into a rectangular grid
    > sq <- ceiling(sqrt(length(methods)+1))
    > par(mfrow=c(ceiling((length(methods)+1)/sq),sq))
    >
    > ## plot and print true graph
    > cat("\n true graph is ------ \n" )
    
     true graph is ------
    > print(A)
     [,1] [,2] [,3]
    [1,] 0.0 0.8 0.0
    [2,] 0.0 0.0 0.8
    [3,] -0.4 0.0 0.0
    > plotGraphEdgeAttr(A, plotStabSelec = FALSE, labels = labels, thres.point = 0,
    + main = "TRUE GRAPH")
    >
    > ## loop over all methods and compute and print/plot estimate
    > for (method in methods){
    + cat("\n result for method", method," ------ \n" )
    +
    + if(!stability){
    + # Option 1): use this estimator as a point estimate
    + Ahat <- getParents(X, environment, method=method, alpha=0.1, pointConf = TRUE)
    + }else{
    + # Option 2): use a stability selection based estimator
    + # with expected number of false positives bounded by EV=2
    + Ahat <- getParentsStable(X, environment, EV=2, method=method, alpha=0.1)
    + }
    +
    + # print and plot estimate (point estimate thresholded if numerical estimates
    + # are returned)
    + print(Ahat)
    + if(!stability)
    + plotGraphEdgeAttr(Ahat, plotStabSelec = FALSE, labels = labels,
    + thres.point = 0.05,
    + main=paste("POINT ESTIMATE FOR METHOD\n", toupper(method)))
    + else
    + plotGraphEdgeAttr(Ahat, plotStabSelec = TRUE, labels = labels,
    + thres.point = 0, main = paste("STABILITY SELECTION
    + ESTIMATE\n FOR METHOD", toupper(method)))
    + }
    
     result for method backShift ------
    3 x 3 sparse Matrix of class "dgCMatrix"
     1 2 3
    1 . 0.802217975 -0.05583784
    2 0.0047767 . 0.78766266
    3 -0.4062989 -0.006501869 .
    
     result for method LINGAM ------
    Error in if (verbose) message("Centering") : argument is of length zero
    Calls: getParents ... runLINGAM -> <Anonymous> -> uselingam -> estLiNGAM -> fastICA
    Execution halted
Flavor: r-patched-solaris-sparc

Version: 0.1.1
Check: tests
Result: ERROR
    Running the tests in ‘tests/testthat.R’ failed.
    Last 13 lines of output:
     8: getParents(X, environment, method = method, alpha = 0.1)
     9: runLINGAM(X, parentsOf, pointConf, setOptions, directed, verbose, result)
     10: pcalg::LINGAM(X, verbose = verbose)
     11: uselingam(X, verbose = verbose)
     12: estLiNGAM(X, only.perm = TRUE, verbose = verbose)
     13: fastICA(X, n.comp = p, tol = fastICA.tol, verbose = if (verbose >= 1) verbose - 1L)
    
     testthat results ================================================================
     OK: 12 SKIPPED: 0 FAILED: 1
     1. Error: Checks output type for LINGAM
    
     Error: testthat unit tests failed
     Execution halted
Flavor: r-patched-solaris-sparc