Skip to content

Releases: SciML/MultiScaleArrays.jl

Allow tuple nodes

08 May 12:58
c7bb8fe
Compare
Choose a tag to compare
Merge pull request #26 from rafaqz/loose_typed_nodes

More types for nodes

Remove upper bound of syntax change

27 Jan 02:36
Compare
Choose a tag to compare
v0.6.0

update for diffeq syntax changes

Broadcast overloads and handling of extra fields

26 Nov 14:13
Compare
Choose a tag to compare
v0.5.0

setup recursive broadcast overloads #10

Update for v0.6

03 Jul 05:30
Compare
Choose a tag to compare
Merge pull request #20 from ScottPJones/spj/node

Additional renaming, use node(s) instead of daughter(s), fix REQUIRE & README.md

Remove Iterators

23 May 20:43
Compare
Choose a tag to compare
v0.3.0

Merge branch 'iter'

General updates + v0.6 depwarn fix

04 May 13:36
Compare
Choose a tag to compare
Merge pull request #11 from JuliaDiffEq/tk/version-specificity

use more specific VERSION cutoff for dot operators

Full DiffEq Compatibility and Improved Performance

15 Feb 04:46
Compare
Choose a tag to compare

There were some performance improvements, the tests are greatly expanded, and full diffeq compatibility. In order to get diffeq compatibility, a dependency on DiffEqBase.jl was needed, but that's lightweight enough (and the compatibility code is really small, just diffeq.jl, and only pertains to size changing and nothing else) that it doesn't deserve a separate package. I think I will take the same approach with whatever NLSolverBase comes up for full compatibility with NLSolve/Optim if necessary. Oh an ForwardDiff.jl works on these now, so that will probably solve the last of the "linking" problems.

Initial Release

07 Feb 15:54
Compare
Choose a tag to compare

This is the initial release of MultiScaleModels.jl. It is a package designed to make modeling at multiple scales easy by allowing the model to be an abstraction of an array, and making this array compatible with standard scientific computing tools like DifferentialEquations.jl and Optim.jl. The indexing structure is very performant for large systems, making it a low-overhead abstraction for building complex models.