From mboxrd@z Thu Jan 1 00:00:00 1970 From: Laura Lazzati Subject: Re: [outreach] Help trying to create R package Date: Wed, 17 Oct 2018 13:38:53 -0300 Message-ID: References: <20181016074412.54757b18@alma-ubu> <20181016082350.75757e6b@alma-ubu> <20181017001851.71543cc9@alma-ubu> <20181017091750.1c408b8d@alma-ubu> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:43608) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gCorN-0005nl-UW for guix-devel@gnu.org; Wed, 17 Oct 2018 12:39:53 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gCorK-0005Yv-1B for guix-devel@gnu.org; Wed, 17 Oct 2018 12:39:49 -0400 Received: from mail-pl1-x644.google.com ([2607:f8b0:4864:20::644]:45861) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1gCorH-0005Bm-V6 for guix-devel@gnu.org; Wed, 17 Oct 2018 12:39:45 -0400 Received: by mail-pl1-x644.google.com with SMTP id y15-v6so12951284plr.12 for ; Wed, 17 Oct 2018 09:39:31 -0700 (PDT) In-Reply-To: List-Id: "Development of GNU Guix and the GNU System distribution." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: guix-devel-bounces+gcggd-guix-devel=m.gmane.org@gnu.org Sender: "Guix-devel" To: =?UTF-8?Q?G=C3=A1bor_Boskovits?= Cc: Guix-devel , Ricardo Wurmus On Wed, Oct 17, 2018 at 11:27 AM Laura Lazzati wrote: > > On Wed, Oct 17, 2018 at 5:21 AM G=C3=A1bor Boskovits wrote: > > > > =C3=9CHi Laura, > > > > > > > > Bj=C3=B6rn H=C3=B6fling ezt =C3=ADr= ta (id=C5=91pont: 2018. okt. 17., Sze 9:17): > >> > >> Hi Laura, > >> > >> On Tue, 16 Oct 2018 22:01:58 -0300 > >> Laura Lazzati wrote: > >> > >> > On Tue, Oct 16, 2018 at 7:19 PM Bj=C3=B6rn H=C3=B6fling > >> > wrote: > >> > > > >> > > Hi Laura, > >> > Hi! I had not answered yet because I thought I would be too verbose. > >> > If so, please do a Laura -v instead of Laura -vvvvvvv. > >> > Thanks for writing :) it encourages me a lot. > >> > >> You're welcome. It's nice to hear from your progress. Verbosity-level > >> is fine :-) > Thanks :) Hope you enjoy my nerd jokes > > >> [..] > >> > >> > > Here is another one: > >> > > > >> > > guix package -s "^r-" | recsel -p name| grep '\S' | sort >r.txt > >> > > >> > Something I faced after writing, was that in the link one of you > >> > provided: [1]https://guix.mdc-berlin.de/ > >> > I have all the packages that are already available in guix. I like > >> > using the guix package command to practice that too, but i found it > >> > practical to see packages easier. I found out that package aspi > >> > (https://CRAN.R-project.org/package=3Daspi) is not available, and I > >> > would like to ask you if you believe it is a good candidate. > >> > One available package, according to [1] is r-abe > >> > (https://cran.r-project.org/web/packages/abe/index.html) > >> > >> Sorry I have no clue of R at all. Could someone else please answer tha= t > >> question, is r-aspi a good candidate? > >> > >> https://cran.r-project.org/web/packages/aspi/index.html > >> > >> > >> > I wanted to see the template, so I ran: > >> > $ guix import cran abe > >> > And got the template, that matches even the version, and everything > >> > stated in the site. > >> > But then I tried: > >> > $ guix package -i r-abe and got no results. > > > > This happens, as the installed guix doesn't know where to find the pack= age definition. There are several ways around that, for example see the -f = flag of guix package, in the invoking guix package section of the manual. > >> > >> > In fact, after running the command you have just provided, r-abe is > >> > not appearing in the r.txt file. > >> > I can't understand why that is happening. Am I missing something, or > >> > is there maybe an inconsistency between [1] and the guix installed i= n > >> > my foreign distro? > >> > >> Concerning r-abe: I don't know how the list at guix-hpc is created > >> , how often it is updated or if there are extra packages in there. In > >> your case, I suppose you lack behind: > >> > >> There is a commit for r-abe from 2019-09-13, so it is only a month old= . > >> I think you are lacking behind. Have you done a `guix pull`? That's th= e > >> command used to update the package definitions. > >> > >> In case you already did a guix pull recently, and you still don't see > >> it, there could be a problem with your $PATH, that the wrong guix is > >> coming first. Do a `which guix`. > >> Guix should be the one in ~/.config/guix/current/bin/guix > > > > > > Yes, please run guix pull, and check which guix output. > I have to take a look at my .bash_history, but I guess somewhere I ran > guix pull because I recall seing something about the $PATH stuff. I > did not run guix output. It doesn'matter, I save my machine state and > run all that now. You were right. I took a look at my .bashrc file, and I had an: export PATH=3D"/home/laura/.guix-profile/bin${PATH:+:}$PATH" I ran $guix pull, did the PATH export in my CLI, then checked the output of $which guix, that was the same that was mentioned by you, commented that previous line in .bashrc, added export PATH=3D"/home/laura/.config/guix/current/bin${PATH:+:}$PATH" and ran again the command to check the available R packages, made a diff with the previous one, and now at least r-bae appears. > > > >> > >> > >> > Just wanted to install it to run guix edit to check if the suggested > >> > packages were already available or if I needed to package them too, > >> > but I have installed r-ade4, which has both imports and suggests, an= d > >> > I have just run: > >> > $guix import cran ade4 > >> > And noticed that it only states in the template the Import package > >> > (MASS) > >> > >> If I see that right, for R we only add the "Imports" packages, as they > >> are absolutely needed for compiling/running. All "Suggests" can be > >> installed/used just at runtime. Please correct me anyone if I'm wrong, > >> I'm not into R. > >> > >> > >> > Here is my template for aspi (I named it for myself r-aspi.scm). I r= an > >> > guix download from its site, and also guix hash to convert it to > >> > base32: > >> > > >> > (define-public r-aspi > >> > (package > >> > (name "r-aspi") > >> > (version "0.2.0") > >> > (source > >> > (origin > >> > (method url-fetch) > >> > (uri (cran-uri "aspi" version)) > >> > (sha256 > >> > (base32 > >> > "a7izhueiqsdjs2eo7dfyb63cqje7mbqz6ennlyoynxxucbhpdnta")))) > >> > (build-system r-build-system) > >> > (home-page > >> > "https://cran.r-project.org/web/packages/aspi/") > >> > (synopsis > >> > "Analysis of Symmetry of Parasitic Infections") > >> > (description > >> > "Tools for the analysis and visualization of bilateral asymmetry > >> > in parasitic infections.") > >> > (license gpl3+))) > >> > And that's were I ended up today (it is still Tuesday, almost 10pm) > >> > >> I haven't done a full review, but that looks good. Next you need to ge= t > >> the guix sources and prepare a patch. > Thanks, I took a look at most of the R packages in cran site, and > chose it because of two reasons: it has to do with bioinformatics due > to its description, and as it does not mention imports so I found it > easy for a first R package contribution. If it works fine, then I can > go on packaging more complicated R packages. > > >> That's described here: > >> > >> https://www.gnu.org/software/guix/manual/en/html_node/Contributing.htm= l#Contributing > >> > >> > >> Bj=C3=B6rn > > > > > > Best regards, > > g_bor > > Thanks both! > Regards, > Laura Thanks again, time to read how to go on with my contribution :) Laura