It seems the short answer is: yes, base::lengths() works and should be fast If it either doesn't work or is slow, you should report it.
On Tue, Nov 13, 2018 at 3:33 PM Pages, Herve <hpa...@fredhutch.org> wrote: > I'm going to try to provide some details, at the risk to confuse you even > more. > > > In BioC 3.8 / R 3.5 base::lengths() was not just slow on some Bioconductor > objects like IRanges or GRanges, it didn't work: > > > > ir <- IRanges(1:21, 20) > > base::lengths(ir) > Error in getListElement(x, i, ...) : > IRanges objects don't support [[, as.list(), lapply(), or unlist() at > the moment > > > Now it works (in BioC 3.9 / R 3.6): > > > > ir <- IRanges(1:21, 20) > > base::lengths(ir) > [1] 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 > > > The change I announced in my original post is about base::lengths() being > now an S4 generic. A standard S4 generic like BiocGenerics::lengths() or > base::lengths() doesn't do any work beyond dispatching to the right method. > What happened in R 3.6 is that base::lengths() is now an S4 generic. This > means that it knows how to take care of dispatching to the right method. > The various "lengths" methods defined in Bioconductor are the ones doing > the real work and they are fast. They have not changed and their speed > should not be affected by this change in base::lengths(). > > > Furthermore: in R 3.6, even though you can't see it, there is a default > "lengths" method defined in base (it's used as a fallback when no other > method applies). In my 1st code chunk above (i.e. when base::lengths() was > not a generic), I'm calling directly this default method (note that this > was the default method of the BiocGenerics::lengths() generic). This > default method is indeed slow or doesn't work on Bioconductor objects. By > calling BiocGenerics::lengths() (in BioC 3.8) or base::lengths() (in BioC > 3.9), dispatch selects the fast method so nothing has changed from that > perspective. > > > Hope this helps, > > H. > > > On 11/13/18 08:59, Håkon Tjeldnes wrote: > Just to be sure, you are saying that now base::lengths is as fast as > BiocGenerics::lengths was ? Else we would still need the Bioc version. > Because the old base implementation was terrible, I will try to verify this > if you are unsure. > ________________________________ > Fra: Michael Lawrence <lawrence.mich...@gene.com><mailto: > lawrence.mich...@gene.com> > Sendt: fredag 2. november 2018 23.50 > Til: hauken_hey...@hotmail.com<mailto:hauken_hey...@hotmail.com> > Kopi: bioc-devel; Hervé Pagès > Emne: Re: [Bioc-devel] Error : object 'lengths' is not exported by > 'namespace:BiocGenerics' > > Yes, please follow Hervé's instructions. > > On Fri, Nov 2, 2018 at 3:41 PM Håkon Tjeldnes <hauken_hey...@hotmail.com > <mailto:hauken_hey...@hotmail.com>> wrote: > Just a quick question, in 3.8 we used biogenerics::lengths for ORFik, > since it was much faster than base, for our dataset 1 second vs 30 minutes. > We should change to S4 version of lengths now ? > > Get Outlook for Android<https://aka.ms/ghei36< > https://urldefense.proofpoint.com/v2/url?u=https-3A__eur02.safelinks.protection.outlook.com_-3Furl-3Dhttps-253A-252F-252Faka.ms-252Fghei36-26data-3D02-257C01-257C-257Ce617c34c45584b70cf7008d6411590f9-257C84df9e7fe9f640afb435aaaaaaaaaaaa-257C1-257C0-257C636767958210704172-26sdata-3DUl3jkwHfZdrwQjkzzBQTKrXdQ5NXQ1vWRrx7vojZEAw-253D-26reserved-3D0&d=DwMFAw&c=eRAMFD45gAfqt84VtBcfhQ&r=BK7q3XeAvimeWdGbWY_wJYbW0WYiZvSXAJJKaaPhzWA&m=VkR31AqDi_FLujiTJm39v-jaLNkPJov4_ZlFjFit9UI&s=C8hC46CIc0bUw9WUxrvaekaCsN1NAlLcyxiuFKgRjCs&e= > >> > > ________________________________ > From: Bioc-devel <bioc-devel-boun...@r-project.org<mailto: > bioc-devel-boun...@r-project.org>> on behalf of Pages, Herve < > hpa...@fredhutch.org<mailto:hpa...@fredhutch.org>> > Sent: Friday, November 2, 2018 6:08:13 PM > To: bioc-devel@r-project.org<mailto:bioc-devel@r-project.org> > Subject: [Bioc-devel] Error : object 'lengths' is not exported by > 'namespace:BiocGenerics' > > Hi developers, > > Starting with R 3.6, base::lengths() can now be treated as an S4 generic > function (like base::length() or base::names()) so there is no need for > us to define our own lengths() generic function in BiocGenerics. > Therefore we've removed the lengths() generic from BiocGenerics (in > version 0.29.1). Note that you can see the S4 generic associated with > base::lengths() by calling getGeneric("lengths") in a fresh R session: > > > getGeneric("lengths") > standardGeneric for "lengths" defined from package "base" > > function (x, use.names = TRUE) > standardGeneric("lengths") > <bytecode: 0x3548c90> > <environment: 0x3540928> > Methods may be defined for arguments: x > Use showMethods("lengths") for currently available ones. > > This used to return NULL in R < 3.6. > > Because of these changes, packages that currently explicitly import > lengths or a lengths method in their NAMESPACE are failing to install > with an error that looks like this: > > ** byte-compile and prepare package for lazy loading > Error : object 'lengths' is not exported by 'namespace:BiocGenerics' > ERROR: lazy loading failed for package 'BubbleTree' > > The packages currently affected by this are: BubbleTree, csaw, > GA4GHclient, matter, NADfinder, and SomaticSignatures. The fix is to > simply remove any explicit import of lengths from the NAMESPACE. > > Note that many other software packages currently fail to install in BioC > 3.9 because they depend on one of the above packages. > > Please let us know if you have questions or need assistance with this. > > H. > > -- > Hervé Pagès > > Program in Computational Biology > Division of Public Health Sciences > Fred Hutchinson Cancer Research Center > 1100 Fairview Ave. N, M1-B514 > P.O. Box 19024 > Seattle, WA 98109-1024 > > E-mail: hpa...@fredhutch.org<mailto:hpa...@fredhutch.org> > Phone: (206) 667-5791 > Fax: (206) 667-1319 > > _______________________________________________ > Bioc-devel@r-project.org<mailto:Bioc-devel@r-project.org> mailing list > > https://nam05.safelinks.protection.outlook.com/?url=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fbioc-devel&data=02%7C01%7C%7C7c4bb0f70b934115a93708d6410fbb9d%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636767933158117589&sdata=0UfC1T6Q52n1V%2FcKb6vR4RT0h5m0DXuWv4C6EwJ2YCA%3D&reserved=0 > < > https://urldefense.proofpoint.com/v2/url?u=https-3A__eur02.safelinks.protection.outlook.com_-3Furl-3Dhttps-253A-252F-252Fstat.ethz.ch-252Fmailman-252Flistinfo-252Fbioc-2Ddevel-26data-3D02-257C01-257C-257Ce617c34c45584b70cf7008d6411590f9-257C84df9e7fe9f640afb435aaaaaaaaaaaa-257C1-257C0-257C636767958210704172-26sdata-3DnWJ1xrPRmDExBtqAQt4QnYojC8btTQCDKbNHNtGXEKg-253D-26reserved-3D0&d=DwMFAw&c=eRAMFD45gAfqt84VtBcfhQ&r=BK7q3XeAvimeWdGbWY_wJYbW0WYiZvSXAJJKaaPhzWA&m=VkR31AqDi_FLujiTJm39v-jaLNkPJov4_ZlFjFit9UI&s=X1qstfX2NwHi1UEDdonTJAg-XpXfkWfEjftcE9o8S3c&e= > > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioc-devel@r-project.org<mailto:Bioc-devel@r-project.org> mailing list > https://stat.ethz.ch/mailman/listinfo/bioc-devel< > https://urldefense.proofpoint.com/v2/url?u=https-3A__eur02.safelinks.protection.outlook.com_-3Furl-3Dhttps-253A-252F-252Fstat.ethz.ch-252Fmailman-252Flistinfo-252Fbioc-2Ddevel-26data-3D02-257C01-257C-257Ce617c34c45584b70cf7008d6411590f9-257C84df9e7fe9f640afb435aaaaaaaaaaaa-257C1-257C0-257C636767958210704172-26sdata-3DnWJ1xrPRmDExBtqAQt4QnYojC8btTQCDKbNHNtGXEKg-253D-26reserved-3D0&d=DwMFAw&c=eRAMFD45gAfqt84VtBcfhQ&r=BK7q3XeAvimeWdGbWY_wJYbW0WYiZvSXAJJKaaPhzWA&m=VkR31AqDi_FLujiTJm39v-jaLNkPJov4_ZlFjFit9UI&s=X1qstfX2NwHi1UEDdonTJAg-XpXfkWfEjftcE9o8S3c&e= > > > > -- > Hervé Pagès > > Program in Computational Biology > Division of Public Health Sciences > Fred Hutchinson Cancer Research Center > 1100 Fairview Ave. N, M1-B514 > P.O. Box 19024 > Seattle, WA 98109-1024 > > E-mail: hpa...@fredhutch.org<mailto:hpa...@fredhutch.org> > Phone: (206) 667-5791 > Fax: (206) 667-1319 > > > [[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
