[Bioc-devel] Experimental data package or ExperimentHub

dear Valerie,

This is a simple R question, but I can't seem to find the answer. I'm
assuming ExperimentHub wants object created by save() and not
saveRDS().

In make-data.R, I have a loop where I programmatically make a
GAlignmentPairs object for each of 4 samples. I'm then using assign()
to assign the value to an appropriate name, e.g.

assign(sample.name[i], x)

where x is a GAlignmentPairs object I've made.

I'm having trouble though saving this to a file. My first try was:

save(sample.name[i], file=paste0(sample.name[i],".rda"))

But this gives:

Error in save(sample.name[i], file = ...
  object 'sample.name[i]' not found

I also tried get(), but it's also giving the "not found" error.

save(get(sample.name[i]), file=paste0(sample.name[i],".rda"))


thanks
Mike

On Thu, Jun 30, 2016 at 6:41 PM, Obenchain, Valerie
<Valerie.Obenchain at roswellpark.org> wrote:

Hi Mike,

Yes, let's go the ExperimentHub route. The focus is on detailed
documentation / real use cases so the vignette you describe sounds like
a good fit.

Essentially you'll put a package together that is very similar to an
experimental data package with a few extra files described in the
ExperimentHubData vignette:


http://www.bioconductor.org/packages/3.4/bioc/vignettes/ExperimentHubData/inst/doc/ExperimentHubData.html

Let me know when you've got the data objects ready and I'll put them in
S3. Once that's done we'll test the package functions and put the
package through a review on the new tracker.

Val



On 06/30/2016 09:27 AM, Michael Love wrote:

I want to generate an experimental data package with a small subset of
alignments. The end goal is a software vignette, where the user would
specify BAM files. I'd like to try out the ExperimentHub submission
process, which is for R data objects. Does it sound reasonable if I
were to have an initial code chunk that loads GAlignments and writes
out BAM files using rtracklayer::export to then show the use of
software with BAM files? There would be so few alignments that it
wouldn't be a large computational or storage expense. Or should I
submit a standard Bioc experimental data package with files in
inst/extdata?

This email message may contain legally privileged and/or confidential information.  If you are not the intended recipient(s), or the employee or agent responsible for the delivery of this message to the intended recipient(s), you are hereby notified that any disclosure, copying, distribution, or use of this email message is prohibited.  If you have received this message in error, please notify the sender immediately by e-mail and delete this email message from your computer. Thank you.

Hi,

There may be more elegant ways but here is one solution.

sample.name <- c("ABC_study_provider", "XYZ_study_provider")
lapply(sample.name,
    function(xx) {
        filename <- paste0(xx, ".rda")
        ## make the GAlignmentPairs
        ## ...
        xx <- GAlignmentPairs()
        save(xx, file=filename)
    })

Val


On 07/14/2016 05:40 AM, Michael Love wrote:

dear Valerie,

This is a simple R question, but I can't seem to find the answer. I'm
assuming ExperimentHub wants object created by save() and not
saveRDS().

In make-data.R, I have a loop where I programmatically make a
GAlignmentPairs object for each of 4 samples. I'm then using assign()
to assign the value to an appropriate name, e.g.

assign(sample.name[i], x)

where x is a GAlignmentPairs object I've made.

I'm having trouble though saving this to a file. My first try was:

save(sample.name[i], file=paste0(sample.name[i],".rda"))

But this gives:

Error in save(sample.name[i], file = ...
  object 'sample.name[i]' not found

I also tried get(), but it's also giving the "not found" error.

save(get(sample.name[i]), file=paste0(sample.name[i],".rda"))


thanks
Mike

On Thu, Jun 30, 2016 at 6:41 PM, Obenchain, Valerie
<Valerie.Obenchain at roswellpark.org> wrote:

Hi Mike,

Yes, let's go the ExperimentHub route. The focus is on detailed
documentation / real use cases so the vignette you describe sounds like
a good fit.

Essentially you'll put a package together that is very similar to an
experimental data package with a few extra files described in the
ExperimentHubData vignette:

http://www.bioconductor.org/packages/3.4/bioc/vignettes/ExperimentHubData/inst/doc/ExperimentHubData.html

Let me know when you've got the data objects ready and I'll put them in
S3. Once that's done we'll test the package functions and put the
package through a review on the new tracker.

Val



On 06/30/2016 09:27 AM, Michael Love wrote:

I want to generate an experimental data package with a small subset of
alignments. The end goal is a software vignette, where the user would
specify BAM files. I'd like to try out the ExperimentHub submission
process, which is for R data objects. Does it sound reasonable if I
were to have an initial code chunk that loads GAlignments and writes
out BAM files using rtracklayer::export to then show the use of
software with BAM files? There would be so few alignments that it
wouldn't be a large computational or storage expense. Or should I
submit a standard Bioc experimental data package with files in
inst/extdata?

This email message may contain legally privileged and/or confidential

information.  If you are not the intended recipient(s), or the employee or
agent responsible for the delivery of this message to the intended
recipient(s), you are hereby notified that any disclosure, copying,
distribution, or use of this email message is prohibited.  If you have
received this message in error, please notify the sender immediately by
e-mail and delete this email message from your computer. Thank you.



This email message may contain legally privileged and/or confidential
information.  If you are not the intended recipient(s), or the employee or
agent responsible for the delivery of this message to the intended
recipient(s), you are hereby notified that any disclosure, copying,
distribution, or use of this email message is prohibited.  If you have
received this message in error, please notify the sender immediately by
e-mail and delete this email message from your computer. Thank you.