P-rep trial design fails

[bauchetg]

Expected Behavior

For Bugs:

‘Could not calculate design: Can’t locate object method “has_replicated_accession_no” via package “Moose::Meta::Class::ANON::SERIAL::40" at lib/CXGN/Trial/TrialDesign/Plugin/Prep.pm line 31.

Environment

Steps to Reproduce

[bauchetg]

Doing some test, it might be related to an R version update.
ie, on my local I get: Error: package or namespace load failed for ‘DiGGer’:
package ‘DiGGer’ was installed by an R version with different internals; it needs to be reinstalled for use with this R version. The installed version was defined for R 3.2.3, checking with package authors if an update is available for R 3.6

[bauchetg]

On my local R instance I got around the R 3.6 version problem fixed by fixing fortran libs location, like: ln /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libgfortran.3.dylib /usr/local/lib/libgfortran.3.dylib.

The breedbase issue may relates to recent creation of plugins for each design type. The rial design object may ommit has_replicated_accession_no @lukasmueller ?

[lukasmueller]

Is that on your local system? Does it work on Cassavabase?

…

On Mar 13, 2020, at 12:09 PM, Guillaume Bauchet ***@***.***> wrote: On my local R instance I got around the R 3.6 version problem fixed by fixing fortran libs location, like: ln /Library/Frameworks/R.framework/Versions/3.6/Resources/lib/libgfortran.3.dylib /usr/local/lib/libgfortran.3.dylib. The breedbase issue may relates to recent creation of plugins for each design type. The rial design object may ommit has_replicated_accession_no @lukasmueller ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[lukasmueller]

This could not be reproduced in a local install (I think by Titima?). Need to investigate more.

[titima15]

@lukasmueller No, I haven't checked the p-rep design. The other day, I checked the greenhouse design.

[lukasmueller]

Digger is apparently not open source, so we cannot use it in production I guess?

[aco46]

I guess we agreed to install it without keeping a copy of it in the production.

…

Sent from my iPhone

On Apr 13, 2020, at 10:12 PM, Lukas Mueller ***@***.***> wrote: Digger is apparently not open source, so we cannot use it in production I guess? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

[aco46]

Here is the package description: Package: DiGGer Version: 1.0.1 Type: Package Date: 2016-05-18 Title: DiGGer design generator under correlation and blocking Author: Neil Coombes <neil.coombes@dpi.nsw.gov.au> Maintainer: Neil Coombes <diggerdes@yahoo.com.au> Depends: R.oo, R.methodsS3, MASS Description: DiGGer searches for A-efficient experimental designs under specified blocking and correlation. LazyLoad: yes License: NSW DPI Freeware Licence URL: http://www.austatgen.org/files/software/downloads Built: R 3.2.3; x86_64-pc-linux-gnu; 2016-05-18 02:12:38 UTC; unix

…

On Mon, Apr 13, 2020 at 10:12 PM Lukas Mueller ***@***.***> wrote: Digger is apparently not open source, so we cannot use it in production I guess? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#2834 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACRALFGY2RUE6IMJWXM4N23RMPBBVANCNFSM4LG7RJBA> .

-- *Alex Ogbonna* Ph.D. Student, Department of Plant Breeding and Genetics, College of Agriculture and Life Sciences, Cornell University, Ithaca, NY 14853, USA 607-229-3580 aco46@cornell.edu

[aco46]

I believe this has been fixed. Should this issue be closed?