Update from igraph0 to igraph

[zbjornson]

Tracking ticket.

[zbjornson]

No, that fixes #74.

[michaelmior]

Awesome that this is fixed. Unfortunately the version on Bioconductor mentioned in the getting started page on the wiki still references igraph0 which is no longer available via CRAN. Is there are another source SPADE is available from that has this issue resolved? (besides installing from source locally)

[zbjornson]

Unfortunately not yet -- wanted to do some more testing of 1.6 before pushing to CRAN/Bioconductor.

Which platform are you on? I can post beta binaries to github today or tomorrow.

[michaelmior]

I'm on Linux. But I was actually testing for a friend who is running on Windows. If you can post both that would be fantastic :D Otherwise just Windows will hopefully work. Thanks!

[zbjornson]

Hi Michael: Sorry for the delay. We just updated SPADE in bioconductor. If all goes well, you should be able to download a build from there within the next 16 hours. http://www.bioconductor.org/checkResults/devel/bioc-LATEST/spade/zin1-buildsrc.html http://www.bioconductor.org/packages/2.13/bioc/html/spade.html (The version is erroneously listed as 1.9...)

[michaelmior]

Thanks so much! I'll let you know when I've had a chance to test :)

[michaelmior]

Hmm, both those links are dead. Is it just a delay issue that the packages haven't been processed? I'm not very familiar with Bioconductor.

[zbjornson]

Ugh. Bioconductor is a bit of a mystery to me as well.

Have a look at the getting started wiki page -- I updated it today and should install the latest from bioc. I'll try tomorrow morning as well. Sorry for the trouble!

[zbjornson]

Just got an email saying the build failed... Not worth trying. Sorry -- I'll just post builds tomorrow.

[zbjornson]

Hi Michael, please try this: https://github.com/nolanlab/spade/releases/download/v1.6.0/spade_1.6.1-WIN.zip We're having some confusion with bioconductor over their versioning scheme, so no builds from them for now, but hopefully that will work on Win x86 and x64.

[michaelmior]

Sorry for the delay. I haven't had a chance to try this until now. Seems to be working great though. Thanks for all your help!

[michaelmior]

I spoke too soon. No problems at all with the install and spade gives no errors when it is initially loaded. However, when I try the example. I first get an error that downsampling_samples is an invalid argument. After removing it, I get the following output

Downsampling file: C:\Users\mmior\Downloads\Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs
  Estimated downsampling-I progress:  0% ...
  Estimated downsampling-I progress: 100% ...
Targeting 9048 events for output/Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs.density.fcs
Clustering files...
Upsampling file: output/Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs.density.fcs
Error in loadNamespace(name) : there is no package called ‘igraph0’
In addition: Warning messages:
1: In write.FCS(out_frame, outfilename) :
  'write.FCS' is not fully tested and should be considered as experimental.
2: In write.FCS(out_frame, outfilename) :
  'write.FCS' is not fully tested and should be considered as experimental.
3: In write.FCS(ff, outfilename) :
  'write.FCS' is not fully tested and should be considered as experimental.
4: In write.FCS(out_frame, outfilename) :
  'write.FCS' is not fully tested and should be considered as experimental.

It seems something somewhere is still trying to load igraph0, but I'm not sure what it is. Any ideas appreciated :)

[zbjornson]

Hi Michael, I made a few bug fixes this morning. They should be available via bioc tonight. (See #78 and #79.) http://bioconductor.org/checkResults/2.13/bioc-LATEST/spade/zin1-buildsrc.html look for 1.10.2 Sorry!

[michaelmior]

Great! Thanks for being so on top of this! Although it looks like the build has failed.

[zbjornson]

Sure, sorry for the issues. The failed build was from before the bugfixes went in... the new build should trigger tonight around 9 PM -0700.

[michaelmior]

No problem at all. Thanks for explaining. I'm not familiar with the build process on bioc.

[zbjornson]

Good afternoon, last night's builds were successful. SPADE is available here: http://bioconductor.org/packages/2.13/bioc/html/spade.html

[shenorr]

Hey Zach,

I just tried installing this as well, off of the 1.10.2 bioconductor page for 64 bit Windows 7. As Michael mentions above there still seems to be a call for igraph0 somewhere which is observed as soon as you try running the example, as per below. Michael did not post any follow up from 18 days back so I am not sure this has been resolved. Also, as an (unrelaeted ?) aside, the Wiki page says invoking the library should yield this: library(spade) Loading required package: igraph Scalable Robust Estimators with High Breakdown Point (version 1.3-01)

but what I get is

library('spade') Loading required package: igraph Loading required package: Rclusterpp Loading required package: Rcpp Loading required package: RcppEigen

SPADE.driver("Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs", out_dir="output", cluster_cols=markers, layout=igraph0:::layout.kamada.kawai, k=200, clustering_samples=50000) Downsampling file: Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs Estimated downsampling-I progress: 0% ... Estimated downsampling-I progress: 100% ... Targeting 9048 events for output/Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs.density.fcs Clustering files... Upsampling file: output/Bendall_et_al_Science_2011_Marrow_1_SurfacePanel_Live_CD44pos_Singlets.fcs.density.fcs Error in loadNamespace(name) : there is no package called ‘igraph0’ In addition: Warning messages: 1: In write.FCS(out_frame, outfilename) : 'write.FCS' is not fully tested and should be considered as experimental. 2: In write.FCS(out_frame, outfilename) : 'write.FCS' is not fully tested and should be considered as experimental. 3: In write.FCS(ff, outfilename) : 'write.FCS' is not fully tested and should be considered as experimental. 4: In write.FCS(out_frame, outfilename) : 'write.FCS' is not fully tested and should be considered as experimental.

thanks,

Shai

[michaelmior]

@shenorr I was still experiencing problems last time I tried, but I have no need to work with SPADE right now.

[zbjornson]

Hi Shai,

Looks like you're using an example script from somewhere that still says layout=igraph0:::layout.kamada.kawai instead of just igraph0. I'm trying to find that but not seeing it... Is that the case? If so, could you point me to the example script so I can update it please?

I'll update the wiki, thanks for reminding me.

Thanks, Zach

[shenorr]

Hey Zach Aha! Yes the script appears in the getting started wiki page

Sent by courier pigeon.

On 11 בנוב 2013, at 20:33, Zach Bjornson notifications@github.com wrote:

Hi Shai,

Looks like you're using an example script from somewhere that still says layout=igraph0:::layout.kamada.kawai instead of just igraph0. I'm trying to find that but not seeing it... Is that the case? If so, could you point me to the example script so I can update it please?

I'll update the wiki, thanks for reminding me.

Thanks, Zach

— Reply to this email directly or view it on GitHub.