[Bioc-devel] Issue tracker for Bioconductor
Hi, I am a graduate student at CMU and I am interested in studying scientific software eco-systems such as Bioconductor. I wanted to know if there is a publicly available issue tracker / bug reports for Bioconductor or something that I can gain read-only access to. Thank you all for your help. Best, -Arun [[alternative HTML version deleted]] ___ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel
Re: [Bioc-devel] writeVcf performance
Martin and Val. I re-ran writeVcf on our (G)VCF data (34790518 ranges, 24 geno fields) with profiling enabled. The results of summaryRprof for that run are attached, though for a variety of reasons they are pretty misleading. It took over an hour to write (3700+seconds), so it's definitely a bottleneck when the data get very large, even if it isn't for smaller data. Michael and I both think the culprit is all the pasting and cbinding that is going on, and more to the point, that memory for an internal representation to be written out is allocated at all. Streaming across the object, looping by rows and writing directly to file (e.g. from C) should be blisteringly fast in comparison. ~G On Tue, Aug 26, 2014 at 11:57 AM, Michael Lawrence micha...@gene.com wrote: Gabe is still testing/profiling, but we'll send something randomized along eventually. On Tue, Aug 26, 2014 at 11:15 AM, Martin Morgan mtmor...@fhcrc.org wrote: I didn't see in the original thread a reproducible (simulated, I guess) example, to be explicit about what the problem is?? Martin On 08/26/2014 10:47 AM, Michael Lawrence wrote: My understanding is that the heap optimization provided marginal gains, and that we need to think harder about how to optimize the all of the string manipulation in writeVcf. We either need to reduce it or reduce its overhead (i.e., the CHARSXP allocation). Gabe is doing more tests. On Tue, Aug 26, 2014 at 9:43 AM, Valerie Obenchain voben...@fhcrc.org wrote: Hi Gabe, Martin responded, and so did Michael, https://stat.ethz.ch/pipermail/bioc-devel/2014-August/006082.html It sounded like Michael was ok with working with/around heap initialization. Michael, is that right or should we still consider this on the table? Val On 08/26/2014 09:34 AM, Gabe Becker wrote: Val, Has there been any movement on this? This remains a substantial bottleneck for us when writing very large VCF files (e.g. variants+genotypes for whole genome NGS samples). I was able to see a ~25% speedup with 4 cores and an optimal speedup of ~2x with 10-12 cores for a VCF with 500k rows using a very naive parallelization strategy and no other changes. I suspect this could be improved on quite a bit, or possibly made irrelevant with judicious use of serial C code. Did you and Martin make any plans regarding optimizing writeVcf? Best ~G On Tue, Aug 5, 2014 at 2:33 PM, Valerie Obenchain voben...@fhcrc.org mailto:voben...@fhcrc.org wrote: Hi Michael, I'm interested in working on this. I'll discuss with Martin next week when we're both back in the office. Val On 08/05/14 07:46, Michael Lawrence wrote: Hi guys (Val, Martin, Herve): Anyone have an itch for optimization? The writeVcf function is currently a bottleneck in our WGS genotyping pipeline. For a typical 50 million row gVCF, it was taking 2.25 hours prior to yesterday's improvements (pasteCollapseRows) that brought it down to about 1 hour, which is still too long by my standards ( 0). Only takes 3 minutes to call the genotypes (and associated likelihoods etc) from the variant calls (using 80 cores and 450 GB RAM on one node), so the output is an issue. Profiling suggests that the running time scales non-linearly in the number of rows. Digging a little deeper, it seems to be something with R's string/memory allocation. Below, pasting 1 million strings takes 6 seconds, but 10 million strings takes over 2 minutes. It gets way worse with 50 million. I suspect it has something to do with R's string hash table. set.seed(1000) end - sample(1e8, 1e6) system.time(paste0(END, =, end)) user system elapsed 6.396 0.028 6.420 end - sample(1e8, 1e7) system.time(paste0(END, =, end)) user system elapsed 134.714 0.352 134.978 Indeed, even this takes a long time (in a fresh session): set.seed(1000) end - sample(1e8, 1e6) end - sample(1e8, 1e7) system.time(as.character(end)) user system elapsed 57.224 0.156 57.366 But running it a second time is faster (about what one would expect?): system.time(levels - as.character(end)) user system elapsed 23.582 0.021 23.589 I did some simple profiling of R to find that the resizing of the string hash table is not a significant component of the time. So maybe something to do with the R heap/gc? No time right now to go deeper. But I know Martin likes this sort of thing ;) Michael [[alternative HTML
Re: [Bioc-devel] Issue tracker for Bioconductor
Hi Arun, - Original Message - From: Arun Kalyanasundaram arunk...@cs.cmu.edu To: bioc-devel@r-project.org Sent: Wednesday, August 27, 2014 10:01:44 AM Subject: [Bioc-devel] Issue tracker for Bioconductor Hi, I am a graduate student at CMU and I am interested in studying scientific software eco-systems such as Bioconductor. I wanted to know if there is a publicly available issue tracker / bug reports for Bioconductor or something that I can gain read-only access to. There is no central tracker. Sometimes bugs are discussed on this mailing list. Some packages have issue trackers, mostly in github. These packages *should* list the issue tracker URL in the package DESCRIPTION field (but don't always). Most packages that are in github as well as svn are listed here: https://gitsvn.bioconductor.org/list_bridges Dan Thank you all for your help. Best, -Arun [[alternative HTML version deleted]] ___ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel ___ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel
Re: [Bioc-devel] writeVcf performance
The profiling I attached in my previous email is for 24 geno fields, as I said, but our typical usecase involves only ~4-6 fields, and is faster but still on the order of dozens of minutes. Sorry for the confusion. ~G On Wed, Aug 27, 2014 at 11:45 AM, Gabe Becker becke...@gene.com wrote: Martin and Val. I re-ran writeVcf on our (G)VCF data (34790518 ranges, 24 geno fields) with profiling enabled. The results of summaryRprof for that run are attached, though for a variety of reasons they are pretty misleading. It took over an hour to write (3700+seconds), so it's definitely a bottleneck when the data get very large, even if it isn't for smaller data. Michael and I both think the culprit is all the pasting and cbinding that is going on, and more to the point, that memory for an internal representation to be written out is allocated at all. Streaming across the object, looping by rows and writing directly to file (e.g. from C) should be blisteringly fast in comparison. ~G On Tue, Aug 26, 2014 at 11:57 AM, Michael Lawrence micha...@gene.com wrote: Gabe is still testing/profiling, but we'll send something randomized along eventually. On Tue, Aug 26, 2014 at 11:15 AM, Martin Morgan mtmor...@fhcrc.org wrote: I didn't see in the original thread a reproducible (simulated, I guess) example, to be explicit about what the problem is?? Martin On 08/26/2014 10:47 AM, Michael Lawrence wrote: My understanding is that the heap optimization provided marginal gains, and that we need to think harder about how to optimize the all of the string manipulation in writeVcf. We either need to reduce it or reduce its overhead (i.e., the CHARSXP allocation). Gabe is doing more tests. On Tue, Aug 26, 2014 at 9:43 AM, Valerie Obenchain voben...@fhcrc.org wrote: Hi Gabe, Martin responded, and so did Michael, https://stat.ethz.ch/pipermail/bioc-devel/2014-August/006082.html It sounded like Michael was ok with working with/around heap initialization. Michael, is that right or should we still consider this on the table? Val On 08/26/2014 09:34 AM, Gabe Becker wrote: Val, Has there been any movement on this? This remains a substantial bottleneck for us when writing very large VCF files (e.g. variants+genotypes for whole genome NGS samples). I was able to see a ~25% speedup with 4 cores and an optimal speedup of ~2x with 10-12 cores for a VCF with 500k rows using a very naive parallelization strategy and no other changes. I suspect this could be improved on quite a bit, or possibly made irrelevant with judicious use of serial C code. Did you and Martin make any plans regarding optimizing writeVcf? Best ~G On Tue, Aug 5, 2014 at 2:33 PM, Valerie Obenchain voben...@fhcrc.org mailto:voben...@fhcrc.org wrote: Hi Michael, I'm interested in working on this. I'll discuss with Martin next week when we're both back in the office. Val On 08/05/14 07:46, Michael Lawrence wrote: Hi guys (Val, Martin, Herve): Anyone have an itch for optimization? The writeVcf function is currently a bottleneck in our WGS genotyping pipeline. For a typical 50 million row gVCF, it was taking 2.25 hours prior to yesterday's improvements (pasteCollapseRows) that brought it down to about 1 hour, which is still too long by my standards ( 0). Only takes 3 minutes to call the genotypes (and associated likelihoods etc) from the variant calls (using 80 cores and 450 GB RAM on one node), so the output is an issue. Profiling suggests that the running time scales non-linearly in the number of rows. Digging a little deeper, it seems to be something with R's string/memory allocation. Below, pasting 1 million strings takes 6 seconds, but 10 million strings takes over 2 minutes. It gets way worse with 50 million. I suspect it has something to do with R's string hash table. set.seed(1000) end - sample(1e8, 1e6) system.time(paste0(END, =, end)) user system elapsed 6.396 0.028 6.420 end - sample(1e8, 1e7) system.time(paste0(END, =, end)) user system elapsed 134.714 0.352 134.978 Indeed, even this takes a long time (in a fresh session): set.seed(1000) end - sample(1e8, 1e6) end - sample(1e8, 1e7) system.time(as.character(end)) user system elapsed 57.224 0.156 57.366 But running it a second time is faster (about what one would expect?): system.time(levels - as.character(end)) user system elapsed 23.582 0.021 23.589 I did some simple profiling of R to find that the resizing of
Re: [Bioc-devel] Issue tracker for Bioconductor
Hi, Arun. There is not such a system that covers the entire Bioconductor project. Since packages are largely contributed by diverse developers, there are many disparate bug tracking systems in use (and many packages with no formal bug tracking). There is a facility in R to allow package authors to specify a bug reporting mechanism. Some details are available on this page: http://stat.ethz.ch/R-manual/R-devel/library/utils/html/bug.report.html Several packages in Bioconductor supply bug reporting links, so you could look into those that do. Sean On Wed, Aug 27, 2014 at 1:01 PM, Arun Kalyanasundaram arunk...@cs.cmu.edu wrote: Hi, I am a graduate student at CMU and I am interested in studying scientific software eco-systems such as Bioconductor. I wanted to know if there is a publicly available issue tracker / bug reports for Bioconductor or something that I can gain read-only access to. Thank you all for your help. Best, -Arun [[alternative HTML version deleted]] ___ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel [[alternative HTML version deleted]] ___ Bioc-devel@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/bioc-devel