Skip to content

Make Turing.jl 0.6 compatible #293

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
xukai92 opened this issue Jun 24, 2017 · 10 comments
Closed

Make Turing.jl 0.6 compatible #293

xukai92 opened this issue Jun 24, 2017 · 10 comments
Assignees
@xukai92
Milestone
Release 0.3.1

Comments

@xukai92
Copy link
Member

xukai92 commented Jun 24, 2017 •
edited
Loading

Typing system is kind of changed - would need some refactors.

Related issue: #173

I would suggest that we fix #177 and also put 0.6 support in our 0.3 release?
@xukai92 xukai92 mentioned this issue Jun 24, 2017
Closed
@yebai yebai added this to the Release 0.4 milestone Jul 26, 2017
@ChrisRackauckas ChrisRackauckas mentioned this issue Aug 23, 2017
Closed
@ChrisRackauckas
Copy link
Collaborator

When is this planned for?

@xukai92 xukai92 mentioned this issue Sep 21, 2017
Closed
@yebai
Copy link
Member

yebai commented Sep 25, 2017

@ChrisRackauckas This is already planed in 0.3.1, which we plan to fix the adaption issue (#324) and Julia 0.6 compatibility.

@emilemathieu
Copy link
Collaborator

Closed the duplicate issue #330

@emilemathieu emilemathieu mentioned this issue Nov 10, 2017
Closed
@yebai yebai modified the milestones: Release 0.4, Release 0.3.1 Nov 14, 2017
yebai added a commit that referenced this issue Nov 16, 2017
@yebai
Solve compatibility with Julia 0.6 (#341, #330, #293)
3f788de 
* Initial support for infix ~ (#173).

* seperate model wrapper call in sample.jl test

* update ForwardDiff.Dual signature

* sample.jl test passed

* fix some tests for Julia 0.6

* remove typealias for 0.6

* make some functions 0.6-ish

* make abstract 0.6-ish

* change some decpreated functions

* Update .travis.yml

* Update appveyor.yml

* Update appveyor.yml

* fix test

* Deprecations on package loading fixed

* fix deprecations

* implement callbacks for inner function

* fix model type bug

* Fix type

* update Dual in benchmark

* update Dual constructor

* Bump up required Julia version to 0.6

* Disable depreciated warning messages for `consume/produce`.

* Remove duplicate definition of produce.

* fix floor, tanh, abs, log

* fix logpdf warning and bug

* fix vec assume init

* Travis: Allow `Benchmarking` test to fail.
@yebai
Copy link
Member

yebai commented Nov 16, 2017

cc @xukai92 @emilemathieu @ae-foster @ChrisRackauckas

We have fixed all known compatibility issues of Turing with Julia 0.6. The master branch of Turing now supports (requires) Julia 0.6.

@yebai yebai closed this as completed Nov 16, 2017
@yebai yebai reopened this Nov 16, 2017
@yebai yebai closed this as completed Nov 16, 2017
@ChrisRackauckas
Copy link
Collaborator

Great thanks! When do you plan on tagging?

@yebai
Copy link
Member

yebai commented Nov 16, 2017 •
edited
Loading

There are some remaining issues with the NUTS sampler, mostly related to robust adaption (see https://github.com/yebai/Turing.jl/milestone/5). I planned to fix those issues altogether in 0.3.1. But perhaps we can move these issues to 0.3.2 and make a new tagging now. What do you think?

cc @xukai92

@ChrisRackauckas
Copy link
Collaborator

I would prefer a quicker tag, and more can always be added later of course. Reviewing tags in METADATA is easier when you have less per tag.

@yebai
Copy link
Member

yebai commented Nov 17, 2017

I see - will do.

emilemathieu pushed a commit that referenced this issue Nov 17, 2017
@yebai @emilemathieu
Solve compatibility with Julia 0.6 (#341, #330, #293)
29e6edf 
* Initial support for infix ~ (#173).

* seperate model wrapper call in sample.jl test

* update ForwardDiff.Dual signature

* sample.jl test passed

* fix some tests for Julia 0.6

* remove typealias for 0.6

* make some functions 0.6-ish

* make abstract 0.6-ish

* change some decpreated functions

* Update .travis.yml

* Update appveyor.yml

* Update appveyor.yml

* fix test

* Deprecations on package loading fixed

* fix deprecations

* implement callbacks for inner function

* fix model type bug

* Fix type

* update Dual in benchmark

* update Dual constructor

* Bump up required Julia version to 0.6

* Disable depreciated warning messages for `consume/produce`.

* Remove duplicate definition of produce.

* fix floor, tanh, abs, log

* fix logpdf warning and bug

* fix vec assume init

* Travis: Allow `Benchmarking` test to fail.
@mkborregaard
Copy link

Would it maybe be an idea to add 0.6 to the travis tracker on the readme.md now this works?

@xukai92
Copy link
Member Author

xukai92 commented Jan 9, 2018

I've added the badge of 0.6 though I think the compilation doesn't pass because of the Stan dependency. For this reason I added a News sections explicitly saying that Turing.jl is 0.6 compatible.

yebai added a commit that referenced this issue Sep 18, 2018
@yebai
Solve compatibility with Julia 0.6 (#341, #330, #293)
28c9f89 
* Initial support for infix ~ (#173).

* seperate model wrapper call in sample.jl test

* update ForwardDiff.Dual signature

* sample.jl test passed

* fix some tests for Julia 0.6

* remove typealias for 0.6

* make some functions 0.6-ish

* make abstract 0.6-ish

* change some decpreated functions

* Update .travis.yml

* Update appveyor.yml

* Update appveyor.yml

* fix test

* Deprecations on package loading fixed

* fix deprecations

* implement callbacks for inner function

* fix model type bug

* Fix type

* update Dual in benchmark

* update Dual constructor

* Bump up required Julia version to 0.6

* Disable depreciated warning messages for `consume/produce`.

* Remove duplicate definition of produce.

* fix floor, tanh, abs, log

* fix logpdf warning and bug

* fix vec assume init

* Travis: Allow `Benchmarking` test to fail.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@xukai92 xukai92

Labels
None yet
Projects
None yet
Milestone
Release 0.3.1
Development

No branches or pull requests
5 participants
@ChrisRackauckas @yebai @xukai92 @emilemathieu @mkborregaard