Yep, Tony saved the world (for Windows users, at least)! Thanks, David
From: [email protected] [mailto:[email protected]] On Behalf Of Andreas Noack Sent: Tuesday, July 19, 2016 7:04 AM To: [email protected] Subject: Re: [julia-stats] Distributions/Rmath, DataFrames/(NullableArrays,CategoricalArrays) David, I just tried on Windows and it seems that we are also all set there now thanks to Tony. On Mon, Jul 18, 2016 at 2:37 PM, Douglas Bates <[email protected] <mailto:[email protected]> > wrote: Problem goes away when updating to latest master. Thanks Andreas. On Monday, July 18, 2016 at 1:33:49 PM UTC-5, Douglas Bates wrote: On Linux Pkg.test("Distribuitons") fails at From worker 2: testing Weibull(20.0, 1.0) From worker 2: testing Weibull(1.0, 2.0) From worker 2: testing Weibull(5.0, 2.0) From worker 2: ERROR: LoadError: On worker 3: LoadError: assertion failed: |cov(Xs,2) - cov(g)| <= 0.01 cov(Xs,2) = [1.55864 0.529817; 0.529817 4.52288] cov(g) = [1.5605 0.53; 0.53 4.508] difference = 0.014877712873417437 > 0.01 in test_approx_eq at ./test.jl:843 in test_mixture at /home/bates/.julia/v0.5/Distributions/test/mixture.jl:121 in include_string at ./loading.jl:380 in include_from_node1 at ./loading.jl:429 in #5 at /home/bates/.julia/v0.5/Distributions/test/runtests.jl:44 in #501 at ./multi.jl:1193 in run_work_thunk at ./multi.jl:844 in macro expansion at ./multi.jl:1193 [inlined] The problem seems to be reproducible. That is, it doesn't look like it is caused by an unfortunate set of values from an RNG. On Monday, July 18, 2016 at 1:27:35 PM UTC-5, Andreas Noack wrote: DataFrames should as of now also work on 0.5 (on Mac and Linux) On Mon, Jul 18, 2016 at 2:25 PM, Andreas Noack <[email protected] <mailto:[email protected]> > wrote: Sorry for not being clear. That was to Doug so it's only about Mac and Linux. I'm not up to date with the Windows issues. On Mon, Jul 18, 2016 at 2:20 PM, David Anthoff <[email protected] <mailto:[email protected]> > wrote: You mean it should work on Win? How so? Distributions depends on StatsFun, StatsFun depends on Rmath and Rmath doesn’t work on Windows. So nothing precompiles and nothing can be used on Windows, right? From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Andreas Noack Sent: Monday, July 18, 2016 11:17 AM To: [email protected] <mailto:[email protected]> Subject: Re: [julia-stats] Distributions/Rmath, DataFrames/(NullableArrays,CategoricalArrays) I haven't tagged Distributions yet so you'll have to checkout master. I've just run the tests locally without errors. On Mon, Jul 18, 2016 at 1:57 PM, David Anthoff <[email protected] <mailto:[email protected]> > wrote: But note that everything is broken on Windows until Rmath.jl works on Windows. Rmath was removed from Base last week without Rmath.jl being ready on Windows… From: [email protected] <mailto:[email protected]> [mailto:[email protected]] On Behalf Of Andreas Noack Sent: Monday, July 18, 2016 10:46 AM To: [email protected] <mailto:[email protected]> Subject: Re: [julia-stats] Distributions/Rmath, DataFrames/(NullableArrays,CategoricalArrays) Distributions should work as of this morning. I think the only thing left for DataFrames is to tag a version for Distributions so we should be pretty close as that could be done now. On Mon, Jul 18, 2016 at 1:38 PM, Douglas Bates <[email protected] <mailto:[email protected]> > wrote: I know people are working hard on reconciling changes w.r.t. Rmath split off as a package and the Distributions package, and converting DataFrames to use CategoricalArrays and NullableArrays. I don't want to cause anyone even more work to respond to repeated questions but it would be handy to have an ETA on when DataFrames and Distributions will be able to pass Pkg.test on version 0.5 and/or v0.4. Also are there things other can do to help - reviewing and revising documentation, reviewing and revising tests, etc.? -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected] <mailto:[email protected]> . For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
