[Bioc-devel] S4Vectors::DataFrame, S4Vectors::cbind() and non-elementwise metadata

[Robert Castelo]

hi Hervé,

thanks for your answer, i also think it isn't clear how one would 
automatically combine the object-wise metadata, specially if there were 
metadata entries with the same name. just wanted to make sure that this 
is a design decision and will remain this way in the future.

best regards,

robert.

On 10/25/2017 05:35 PM, Hervé Pagès wrote:
> Hi Robert,
> 
> The object-wise metadata doesn't follow any convention and is typically
> used to keep track of things like the file the object is coming from
> (if any), the date/time it was created, by who, the call that was used
> to create it, plus any other relevant information about the object.
> 
> It's not clear to me how n-ary operations should combine the object-wise
> metadata in general, or that it would make sense to try to propagate
> this information from all the input objects to the output object.
> 
> Note that AFAIK most unary operations on Annotated derivatives (e.g.
> shift()) propagate the object-wise metadata, even though it's not
> clear that this information will still be relevant for the transformed
> object.
> 
> Cheers,
> H.
> 
> 
> On 10/25/2017 07:39 AM, Robert Castelo wrote:
>> hi,
>>
>> i've the following question on how 'S4Vectors::cbind()' works on
>> 'DataFrame' objects.
>>
>> let's say i have the following two toy 'DataFrame' objects, each of them
>> including some elementwise and non-elementwise metadata (i hope i'm
>> using the right terminology!):
>>
>> library(S4Vectors)
>>
>> dtf1 <- DataFrame(col1=1:10)
>> mcols(dtf1) <- DataFrame(ElementwiseMeta="This is column1")
>> metadata(dtf1) <- list(Meta1="Object storing column1")
>>
>> dtf2 <- DataFrame(col2=1:10)
>> mcols(dtf2) <- DataFrame(ElementwiseMeta="This is column2")
>> metadata(dtf2) <- list(Meta2="Object storing column2")
>>
>> now use 'cbind()' to combine the two 'DataFrame' objects:
>>
>> dtf <- cbind(dtf1, dtf2)
>>
>> the elementwise metadata is also combined:
>>
>> mcols(dtf)
>> DataFrame with 2 rows and 1 column
>>    ElementwiseMeta
>>        <character>
>> 1 This is column1
>> 2 This is column2
>>
>> but the non-elementwise metadata is not:
>>
>> metadata(dtf)
>> list()
>>
>> is there a particular reason why the non-elementwise metadata is not
>> also combined?
>>
>> please find my session information below.
>>
>> thanks!!!
>>
>> robert.
>> ps: sessionInfo()
>> R version 3.4.0 (2017-04-21)
>> Platform: x86_64-pc-linux-gnu (64-bit)
>> Running under: CentOS Linux 7 (Core)
>>
>> Matrix products: default
>> BLAS: /opt/R/R-3.4.0_BioCdevel/lib64/R/lib/libRblas.so
>> LAPACK: /opt/R/R-3.4.0_BioCdevel/lib64/R/lib/libRlapack.so
>>
>> locale:
>>   [1] LC_CTYPE=en_US.UTF8       LC_NUMERIC=C
>>   [3] LC_TIME=en_US.UTF8        LC_COLLATE=en_US.UTF8
>>   [5] LC_MONETARY=en_US.UTF8    LC_MESSAGES=en_US.UTF8
>>   [7] LC_PAPER=en_US.UTF8       LC_NAME=C
>>   [9] LC_ADDRESS=C              LC_TELEPHONE=C
>> [11] LC_MEASUREMENT=en_US.UTF8 LC_IDENTIFICATION=C
>>
>> attached base packages:
>> [1] parallel  stats4    stats     graphics  grDevices utils     datasets
>> [8] methods   base
>>
>> other attached packages:
>> [1] S4Vectors_0.15.14   BiocGenerics_0.23.4 colorout_1.1-2
>>
>> loaded via a namespace (and not attached):
>> [1] compiler_3.4.0
>>
>> _______________________________________________
>> Bioc-devel at r-project.org mailing list
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__stat.ethz.ch_mailman_listinfo_bioc-2Ddevel&d=DwICAg&c=eRAMFD45gAfqt84VtBcfhQ&r=BK7q3XeAvimeWdGbWY_wJYbW0WYiZvSXAJJKaaPhzWA&m=A-QAZvhpPY0ytgF8msL4_W8jBMlWkyL7yLBW4QwXFqw&s=DgYYXQl-0a3XcBf-IKJUPox3jKkiPCmqVNQTFAq0XdA&e= 
>>
>>
> 

-- 
Robert Castelo, PhD
Associate Professor
Dept. of Experimental and Health Sciences
Universitat Pompeu Fabra (UPF)
Barcelona Biomedical Research Park (PRBB)
Dr Aiguader 88
E-08003 Barcelona, Spain
telf: +34.933.160.514
fax: +34.933.160.550