[Bioc-devel] unable to push local changes to bioconductor svn repository

2017-05-11 Thread Xu, Zongli (NIH/NIEHS) [E]
Hi, I got a new Linux computer, cloned my package from github, made some changes. I am able to update the changes to github, but can not push it to Bioconductor svn repository. Can someone help me to solve the problem. The following is what I did: git svn clone https://github.com/xuz1/EN

[Bioc-devel] Question about BiocGeneric::order

2017-05-11 Thread Ou, Jianhong
I got error when I try order for Rle object by following codes: library("BiocGenerics") library(XVector) order(Rle(1)) ## Error in match.arg(method) : 'arg' must be of length 1 > sessionInfo() R Under development (unstable) (2017-05-10 r72667) Platform: x86_64-apple-darwin16.5.0 (64-bit) Running

Re: [Bioc-devel] Question about BiocGeneric::order

2017-05-11 Thread Hervé Pagès
Hi Jianhong, I can't reproduce this but I'm using R 3.4.0. You seem to be using Bioc devel (aka BioC 3.6) with R devel. This is not supported. Both, BioC 3.5 (current release) and BioC 3.6 require R 3.4.0. Cheers, H. On 05/11/2017 01:11 PM, Ou, Jianhong wrote: I got error when I try order for

Re: [Bioc-devel] Question about BiocGeneric::order

2017-05-11 Thread Ou, Jianhong
Thank you Hervé, I got that. Good to know that BioC 3.6 require R 3.4.0. Yours Sincerely, Jianhong Ou TEL: 508-856-5379 LRB 608 Bioinformatician of Bioinformatics core at Department of Molecular, Cell and Cancer Biology UMASS Medical School 364 Plantation Street Worcester, MA 01605 Confidenti

Re: [Bioc-devel] Question about BiocGeneric::order

2017-05-11 Thread Michael Lawrence
There is a bug in S4Vectors, but thanks to an R 3.4 bug in the methods package (recently fixed in devel), the bug is masked. So, we should fix S4Vectors. The problem is that order,Rle has a default for the 'method' argument that differs from that of the generic. Since R 3.3, base::order() is smart

Re: [Bioc-devel] Question about BiocGeneric::order

2017-05-11 Thread Hervé Pagès
Hi Michael, On 05/11/2017 03:27 PM, Michael Lawrence wrote: There is a bug in S4Vectors, but thanks to an R 3.4 bug in the methods package (recently fixed in devel), the bug is masked. So, we should fix S4Vectors. The problem is that order,Rle has a default for the 'method' argument that differs

Re: [Bioc-devel] Question about BiocGeneric::order

2017-05-11 Thread Hervé Pagès
On 05/11/2017 06:44 PM, Hervé Pagès wrote: Hi Michael, On 05/11/2017 03:27 PM, Michael Lawrence wrote: There is a bug in S4Vectors, but thanks to an R 3.4 bug in the methods package (recently fixed in devel), the bug is masked. So, we should fix S4Vectors. The problem is that order,Rle has a de