CRAN Package Check Results for Package kernlab

Last updated on 2015-11-20 00:47:32.

Memtest notes: valgrind

Check Details

Version: 0.9-22
Check: examples
Result: ERROR
    Running examples in ‘kernlab-Ex.R’ failed
    The error most likely occurred in:
    
    > ### Name: ksvm
    > ### Title: Support Vector Machines
    > ### Aliases: ksvm ksvm,formula-method ksvm,vector-method ksvm,matrix-method
    > ### ksvm,kernelMatrix-method ksvm,list-method show,ksvm-method
    > ### coef,ksvm-method
    > ### Keywords: methods regression nonlinear classif neural
    >
    > ### ** Examples
    >
    >
    > ## simple example using the spam data set
    > data(spam)
    >
    > ## create test and training set
    > index <- sample(1:dim(spam)[1])
    > spamtrain <- spam[index[1:floor(dim(spam)[1]/2)], ]
    > spamtest <- spam[index[((ceiling(dim(spam)[1]/2)) + 1):dim(spam)[1]], ]
    >
    > ## train a support vector machine
    > filter <- ksvm(type~.,data=spamtrain,kernel="rbfdot",
    + kpar=list(sigma=0.05),C=5,cross=3)
    > filter
    Support Vector Machine object of class "ksvm"
    
    SV type: C-svc (classification)
     parameter : cost C = 5
    
    Gaussian Radial Basis kernel function.
     Hyperparameter : sigma = 0.05
    
    Number of Support Vectors : 922
    
    Objective Function Value : -1073.61
    Training error : 0.017391
    Cross validation error : 0.09348
    >
    > ## predict mail type on the test set
    > mailtype <- predict(filter,spamtest[,-58])
    >
    > ## Check results
    > table(mailtype,spamtest[,58])
    
    mailtype nonspam spam
     nonspam 1349 104
     spam 55 792
    >
    >
    > ## Another example with the famous iris data
    > data(iris)
    >
    > ## Create a kernel function using the build in rbfdot function
    > rbf <- rbfdot(sigma=0.1)
    > rbf
    Gaussian Radial Basis kernel function.
     Hyperparameter : sigma = 0.1
    >
    > ## train a bound constraint support vector machine
    > irismodel <- ksvm(Species~.,data=iris,type="C-bsvc",
    + kernel=rbf,C=10,prob.model=TRUE)
    >
    > irismodel
    Support Vector Machine object of class "ksvm"
    
    SV type: C-bsvc (classification)
     parameter : cost C = 10
    
    Gaussian Radial Basis kernel function.
     Hyperparameter : sigma = 0.1
    
    Number of Support Vectors : 32
    
    Objective Function Value : -5.8442 -3.0652 -136.9786
    Training error : 0.02
    Probability model included.
    >
    > ## get fitted values
    > fitted(irismodel)
     [1] setosa setosa setosa setosa setosa setosa
     [7] setosa setosa setosa setosa setosa setosa
     [13] setosa setosa setosa setosa setosa setosa
     [19] setosa setosa setosa setosa setosa setosa
     [25] setosa setosa setosa setosa setosa setosa
     [31] setosa setosa setosa setosa setosa setosa
     [37] setosa setosa setosa setosa setosa setosa
     [43] setosa setosa setosa setosa setosa setosa
     [49] setosa setosa versicolor versicolor versicolor versicolor
     [55] versicolor versicolor versicolor versicolor versicolor versicolor
     [61] versicolor versicolor versicolor versicolor versicolor versicolor
     [67] versicolor versicolor versicolor versicolor versicolor versicolor
     [73] virginica versicolor versicolor versicolor versicolor virginica
     [79] versicolor versicolor versicolor versicolor versicolor virginica
     [85] versicolor versicolor versicolor versicolor versicolor versicolor
     [91] versicolor versicolor versicolor versicolor versicolor versicolor
     [97] versicolor versicolor versicolor versicolor virginica virginica
    [103] virginica virginica virginica virginica virginica virginica
    [109] virginica virginica virginica virginica virginica virginica
    [115] virginica virginica virginica virginica virginica virginica
    [121] virginica virginica virginica virginica virginica virginica
    [127] virginica virginica virginica virginica virginica virginica
    [133] virginica virginica virginica virginica virginica virginica
    [139] virginica virginica virginica virginica virginica virginica
    [145] virginica virginica virginica virginica virginica virginica
    Levels: setosa versicolor virginica
    >
    > ## Test on the training set with probabilities as output
    > predict(irismodel, iris[,-5], type="probabilities")
     setosa versicolor virginica
     [1,] 0.983846258 0.0093172557 0.006836487
     [2,] 0.978645002 0.0136052278 0.007749770
     [3,] 0.985595610 0.0076995180 0.006704872
     [4,] 0.982113992 0.0101230747 0.007762933
     [5,] 0.984928589 0.0083717855 0.006699626
     [6,] 0.971826655 0.0187151799 0.009458166
     [7,] 0.983692067 0.0085001724 0.007807760
     [8,] 0.982711888 0.0100873934 0.007200719
     [9,] 0.981979606 0.0100189253 0.008001469
     [10,] 0.982355257 0.0105661519 0.007078591
     [11,] 0.979825873 0.0125007307 0.007673396
     [12,] 0.982833202 0.0096232859 0.007543512
     [13,] 0.983558214 0.0096630828 0.006778704
     [14,] 0.988782361 0.0049129924 0.006304647
     [15,] 0.973550163 0.0172190287 0.009230808
     [16,] 0.955485351 0.0315096808 0.013004969
     [17,] 0.977381624 0.0144474484 0.008170928
     [18,] 0.981469167 0.0110723459 0.007458488
     [19,] 0.967226516 0.0221546049 0.010618879
     [20,] 0.981151103 0.0112845970 0.007564300
     [21,] 0.971439252 0.0190617313 0.009499017
     [22,] 0.978727843 0.0130459743 0.008226183
     [23,] 0.988154939 0.0053392397 0.006505821
     [24,] 0.955673058 0.0314567092 0.012870233
     [25,] 0.977557181 0.0134083536 0.009034466
     [26,] 0.970756269 0.0200715043 0.009172227
     [27,] 0.973605496 0.0168919244 0.009502579
     [28,] 0.981406810 0.0112214032 0.007371787
     [29,] 0.981698417 0.0110561685 0.007245414
     [30,] 0.981128093 0.0109116188 0.007960288
     [31,] 0.978128169 0.0136077379 0.008264094
     [32,] 0.966327183 0.0231136774 0.010559140
     [33,] 0.978848454 0.0130561800 0.008095366
     [34,] 0.972843038 0.0180108522 0.009146110
     [35,] 0.978901552 0.0132224703 0.007875978
     [36,] 0.984464471 0.0089774116 0.006558117
     [37,] 0.979050478 0.0130484480 0.007901074
     [38,] 0.986722925 0.0069107848 0.006366291
     [39,] 0.984769854 0.0078553995 0.007374746
     [40,] 0.981576714 0.0110594132 0.007363873
     [41,] 0.983690798 0.0092926974 0.007016504
     [42,] 0.955555569 0.0314899424 0.012954489
     [43,] 0.986087895 0.0066221473 0.007289958
     [44,] 0.964813918 0.0232669357 0.011919147
     [45,] 0.972417749 0.0177609147 0.009821336
     [46,] 0.975610321 0.0157396494 0.008650030
     [47,] 0.981867827 0.0107258304 0.007406343
     [48,] 0.984750095 0.0080707392 0.007179166
     [49,] 0.981256923 0.0113829477 0.007360130
     [50,] 0.982983191 0.0099843630 0.007032446
     [51,] 0.029252548 0.9432104200 0.027537032
     [52,] 0.017040714 0.9574335995 0.025525686
     [53,] 0.017020176 0.8611546378 0.121825186
     [54,] 0.006868112 0.8938155097 0.099316378
     [55,] 0.006423721 0.8683449893 0.125231289
     [56,] 0.005398649 0.9434571513 0.051144199
     [57,] 0.019806843 0.9084043471 0.071788810
     [58,] 0.043000208 0.9383465157 0.018653276
     [59,] 0.009467780 0.9704681774 0.020064043
     [60,] 0.007999976 0.9358774058 0.056122619
     [61,] 0.036784196 0.9007548094 0.062460994
     [62,] 0.008183214 0.9651360624 0.026680723
     [63,] 0.008031775 0.9825588169 0.009409409
     [64,] 0.006137838 0.9049620860 0.088900076
     [65,] 0.015797579 0.9797664145 0.004436006
     [66,] 0.017961262 0.9700646712 0.011974067
     [67,] 0.008236733 0.8923225638 0.099440704
     [68,] 0.009546346 0.9870752888 0.003378365
     [69,] 0.008713214 0.6167995983 0.374487188
     [70,] 0.006872366 0.9844126186 0.008715016
     [71,] 0.013435904 0.4992946437 0.487269452
     [72,] 0.006986019 0.9876023016 0.005411680
     [73,] 0.007755773 0.4123867345 0.579857493
     [74,] 0.005602458 0.9644623579 0.029935184
     [75,] 0.008957570 0.9831986713 0.007843758
     [76,] 0.011983418 0.9729047249 0.015111857
     [77,] 0.009051875 0.8914741144 0.099474011
     [78,] 0.009980970 0.4002409694 0.589778061
     [79,] 0.005995640 0.8916738239 0.102330536
     [80,] 0.017227237 0.9788248594 0.003947903
     [81,] 0.007736746 0.9796577941 0.012605460
     [82,] 0.011303741 0.9819048019 0.006791457
     [83,] 0.007016664 0.9881172203 0.004866115
     [84,] 0.007498820 0.1372280131 0.855273167
     [85,] 0.009482424 0.8697147495 0.120802826
     [86,] 0.031451830 0.9376416179 0.030906553
     [87,] 0.014454203 0.9213480591 0.064197738
     [88,] 0.005342455 0.9225469306 0.072110614
     [89,] 0.013052157 0.9792918529 0.007655990
     [90,] 0.005335193 0.9447675024 0.049897305
     [91,] 0.005444922 0.9349869163 0.059568162
     [92,] 0.007834906 0.9504791047 0.041685989
     [93,] 0.005303167 0.9853804132 0.009316420
     [94,] 0.035158731 0.9460700446 0.018771225
     [95,] 0.005137955 0.9610656311 0.033796414
     [96,] 0.014202706 0.9803778900 0.005419404
     [97,] 0.007889492 0.9799842032 0.012126305
     [98,] 0.007804632 0.9830533604 0.009142008
     [99,] 0.042477484 0.9454068281 0.012115688
    [100,] 0.006314068 0.9803855909 0.013300341
    [101,] 0.007991647 0.0010109304 0.990997422
    [102,] 0.005332873 0.0114868076 0.983180320
    [103,] 0.006409118 0.0030929276 0.990497954
    [104,] 0.006063001 0.0136974481 0.980239551
    [105,] 0.005072606 0.0009385997 0.993988794
    [106,] 0.008713017 0.0022574179 0.989029565
    [107,] 0.012024086 0.0897338046 0.898242109
    [108,] 0.007997014 0.0058757465 0.986127240
    [109,] 0.006810189 0.0056123930 0.987577418
    [110,] 0.013033674 0.0096081804 0.977358146
    [111,] 0.010189594 0.0898858563 0.899924550
    [112,] 0.005596535 0.0113254728 0.983077992
    [113,] 0.006245464 0.0063787840 0.987375752
    [114,] 0.004750465 0.0045486277 0.990700908
    [115,] 0.004392511 0.0006269814 0.994980507
    [116,] 0.007839240 0.0064511667 0.985709594
    [117,] 0.007266075 0.0366560505 0.956077875
    [118,] 0.020632163 0.0324084174 0.946959419
    [119,] 0.015081915 0.0024997039 0.982418381
    [120,] 0.010873209 0.1700223871 0.819104404
    [121,] 0.007034392 0.0033684608 0.989597148
    [122,] 0.006062250 0.0143784108 0.979559339
    [123,] 0.011227172 0.0028504988 0.985922329
    [124,] 0.006991988 0.0916547112 0.901353301
    [125,] 0.008594602 0.0139305109 0.977474887
    [126,] 0.009212634 0.0311918684 0.959595498
    [127,] 0.007466557 0.1531821316 0.839351312
    [128,] 0.009006663 0.2104328651 0.780560472
    [129,] 0.004567903 0.0012580956 0.994174002
    [130,] 0.010834395 0.0998162682 0.889349337
    [131,] 0.008677507 0.0061921326 0.985130360
    [132,] 0.029915312 0.0845832428 0.885501445
    [133,] 0.004417017 0.0006952928 0.994887690
    [134,] 0.007765247 0.4270054764 0.565229277
    [135,] 0.008325895 0.0898943692 0.901779736
    [136,] 0.009732419 0.0035379575 0.986729623
    [137,] 0.010964820 0.0062585086 0.982776671
    [138,] 0.008377115 0.0544356248 0.937187260
    [139,] 0.009313238 0.2649907012 0.725696061
    [140,] 0.007373053 0.0153894226 0.977237524
    [141,] 0.006184894 0.0013085317 0.992506575
    [142,] 0.007904553 0.0122058254 0.979889621
    [143,] 0.005332873 0.0114868076 0.983180320
    [144,] 0.006578351 0.0018115477 0.991610102
    [145,] 0.008166540 0.0021704423 0.989663018
    [146,] 0.006400448 0.0043670753 0.989232477
    [147,] 0.006266524 0.0215252549 0.972208221
    [148,] 0.006936193 0.0247416542 0.968322153
    [149,] 0.012534914 0.0171025695 0.970362516
    [150,] 0.008746406 0.0986049013 0.892648693
    >
    >
    > ## Demo of the plot function
    > x <- rbind(matrix(rnorm(120),,2),matrix(rnorm(120,mean=3),,2))
    > y <- matrix(c(rep(1,60),rep(-1,60)))
    >
    > svp <- ksvm(x,y,type="C-svc")
    > plot(svp,data=x)
    >
    >
    > ### Use kernelMatrix
    > K <- as.kernelMatrix(crossprod(t(x)))
    >
    > svp2 <- ksvm(K, y, type="C-svc")
    >
    > svp2
    Support Vector Machine object of class "ksvm"
    
    SV type: C-svc (classification)
     parameter : cost C = 1
    
    [1] " Kernel matrix used as input."
    
    Number of Support Vectors : 7
    
    Objective Function Value : -4.3822
    Training error : 0.008333
    >
    > # test data
    > xtest <- rbind(matrix(rnorm(20),,2),matrix(rnorm(20,mean=3),,2))
    > # test kernel matrix i.e. inner/kernel product of test data with
    > # Support Vectors
    >
    > Ktest <- as.kernelMatrix(crossprod(t(xtest),t(x[SVindex(svp2), ])))
    >
    > predict(svp2, Ktest)
     [1] 1 1 1 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1
    >
    >
    > #### Use custom kernel
    >
    > k <- function(x,y) {(sum(x*y) +1)*exp(-0.001*sum((x-y)^2))}
    > class(k) <- "kernel"
    >
    > data(promotergene)
    >
    > ## train svm using custom kernel
    > gene <- ksvm(Class~.,data=promotergene[c(1:20, 80:100),],kernel=k,
    + C=5,cross=5)
    >
    > gene
    Support Vector Machine object of class "ksvm"
    
    SV type: C-svc (classification)
     parameter : cost C = 5
    
    
    Number of Support Vectors : 41
    
    Objective Function Value : -0.5191
    Training error : 0
    Cross validation error : 0.169444
    >
    >
    > #### Use text with string kernels
    > data(reuters)
    > is(reuters)
    [1] "list" "vector" "input" "listI" "lpinput" "output"
    > tsv <- ksvm(reuters,rlabels,kernel="stringdot",
    + kpar=list(length=5),cross=3,C=10)
    
     *** caught segfault ***
    address 100d6058, cause 'memory not mapped'
    
    Traceback:
     1: .Call("stringtv", as.character(x[i]), as.character(x[i:length(x)]), as.integer(length(x) - i + 1), as.integer(nchar(x[i])), as.integer(nchar(x[i:length(x)])), as.integer(sktype), as.double(kpar(kernel)$lambda))
     2: kernelMatrix(kernel, x[c(indexes[[i]], indexes[[j]])])
     3: kernelMatrix(kernel, x[c(indexes[[i]], indexes[[j]])])
     4: .local(x, ...)
     5: ksvm(reuters, rlabels, kernel = "stringdot", kpar = list(length = 5), cross = 3, C = 10)
     6: ksvm(reuters, rlabels, kernel = "stringdot", kpar = list(length = 5), cross = 3, C = 10)
    aborting ...
Flavor: r-patched-solaris-sparc