[Bioc-devel] asVCF error coming from normalizeSingleBracketSubscript

[Stephanie M. Gogarten]

x is not always a NumericList - sometimes it's a CharacterList.

On 11/25/13 2:45 PM, Hervé Pagès wrote:
> Hi Val, Stephanie,
>
> I wonder what's the purpose of comparing a NumericList with the
> empty string in the first place. That's always going to return a
> LogicalList filled with FALSEs. So x[x %in% ""] <- NA is always
> going to be a no-op. There might be a deeper issue with this code.
>
> Cheers,
> H.
>
>
> On 11/25/2013 02:08 PM, Valerie Obenchain wrote:
>> Hi Stephanie,
>>
>> The error is thrown from SeqArray:::.info at line 216 in the file and is
>> related to the handing of NA values.
>>
>>    x[x == ""] <- NA
>>
>> Output from the == comparison can contain NAs and therefore can't be
>> used (consistently) in subsetting operations.
>>
>> 'x' is a NumericList.
>>
>> Browse[2]> x
>> NumericList of length 5
>> [[1]] 0.5
>> [[2]] 0.017000000923872
>> [[3]] 0.333000004291534 0.666999995708466
>> [[4]] <NA>
>> [[5]] <NA> <NA>
>>
>> Here we see NAs returned for the NA values,
>>
>> Browse[2]> x == ""
>> LogicalList of length 5
>> [[1]] FALSE
>> [[2]] FALSE
>> [[3]] FALSE FALSE
>> [[4]] <NA>
>> [[5]] <NA> <NA>
>>
>> which fail on subsetting.
>>
>> Browse[2]> x[x == ""]
>> Error in normalizeSingleBracketSubscript(i, x) : subscript contains NAs
>>
>> One solution is use %in% which does not return NAs.
>>
>> Browse[2]> x %in% ""
>> LogicalList of length 5
>> [[1]] FALSE
>> [[2]] FALSE
>> [[3]] FALSE FALSE
>> [[4]] FALSE
>> [[5]] FALSE FALSE
>>
>>
>> Valerie
>>
>>
>> On 11/22/2013 03:11 PM, Stephanie M. Gogarten wrote:
>>> Hi Valerie,
>>>
>>> The asVCF method in SeqArray is failing as of today with a (to me)
>>> mysterious error.  I get it for the test files chr22.vcf.gz, ex2.vcf,
>>> and gl_chr1.vcf in extdata of VariantAnnotation, but not for
>>> SeqArray/extdata/CEU_Exon.vcf.  Do you have any suggestions of where I
>>> might look to figure out where this error is coming from?
>>>
>>> thanks,
>>> Stephanie
>>>
>>>  > vcffile <- system.file("extdata", "ex2.vcf",
>>> package="VariantAnnotation")
>>>  > gdsfile <- tempfile()
>>>  > seqVCF2GDS(vcffile, gdsfile)
>>>  > gdsobj <- seqOpen(gdsfile)
>>>  > options(error=recover)
>>>  > vcfg <- asVCF(gdsobj)
>>> Error in normalizeSingleBracketSubscript(i, x) : subscript contains NAs
>>>
>>> Enter a frame number, or 0 to exit
>>>
>>>   1: asVCF(gdsobj)
>>>   2: asVCF(gdsobj)
>>>   3: .local(x, ...)
>>>   4: VCF(rowData = .rowData(x), colData = .colData(x), exptData =
>>> SimpleList(hea
>>>   5: .info(x, info)
>>>   6: `[<-`(`*tmp*`, x == "", value = NA)
>>>   7: `[<-`(`*tmp*`, x == "", value = NA)
>>>   8: lsubset_List_by_List(x, i, value)
>>>   9: .fast_lsubset_List_by_List(x, i, value)
>>> 10: replaceROWS(unlisted_x, unlisted_i, unlisted_value)
>>> 11: replaceROWS(unlisted_x, unlisted_i, unlisted_value)
>>> 12: extractROWS(setNames(seq_along(x), names(x)), i)
>>> 13: extractROWS(setNames(seq_along(x), names(x)), i)
>>> 14: normalizeSingleBracketSubscript(i, x)
>>>
>>>  > sessionInfo()
>>> R version 3.0.2 (2013-09-25)
>>> Platform: x86_64-apple-darwin10.8.0 (64-bit)
>>>
>>> locale:
>>> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
>>>
>>> attached base packages:
>>> [1] parallel  stats     graphics  grDevices utils     datasets  methods
>>> [8] base
>>>
>>> other attached packages:
>>> [1] VariantAnnotation_1.8.6 Rsamtools_1.14.1        Biostrings_2.30.1
>>> [4] GenomicRanges_1.14.3    XVector_0.2.0           IRanges_1.20.6
>>> [7] BiocGenerics_0.8.0      SeqArray_1.2.0          gdsfmt_1.0.0
>>>
>>> loaded via a namespace (and not attached):
>>>   [1] AnnotationDbi_1.24.0   Biobase_2.22.0         biomaRt_2.18.0
>>>   [4] bitops_1.0-6           BSgenome_1.30.0        DBI_0.2-7
>>>   [7] GenomicFeatures_1.14.2 RCurl_1.95-4.1         RSQLite_0.11.4
>>> [10] rtracklayer_1.22.0     stats4_3.0.2           tools_3.0.2
>>> [13] XML_3.95-0.2           zlibbioc_1.8.0
>>
>> _______________________________________________
>> Bioc-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>