Releases: jump-dev/Convex.jl
Releases · jump-dev/Convex.jl
v0.10.1
Fix method overwrite warning when loading the package (#255) This was caused by providing a default value of `nothing` for the `model` argument in the `Problem` constructor, which ends up creating a two-argument method that is also created further down by virtue of using default values. By simply removing the default value for `model`, we remain at feature parity with the methods prior to the change and we don't overwrite anything.
v0.10.0
Completely drop the DEFAULT_SOLVER logic (#250) - Dropped get_default_solver and set_default_solver - Reorganized tests to not depend on get_default_solver (the giant at-testset for loop can be changed). - Only API change is that if you didn't define a problem with a solver/model then `solve!(p)` throws an argument (e.g., `solve!(p, solver)` is preferred now. NOTE: The existing `solve!(p, solver)` just assigns the problem model anyways with `problem.model = MathProgBase.ConicModel(s)`
v0.9.1
v0.9.0
v0.8.0
v0.7.0
Compatibility with v0.6
This release introduces breaking changes for compatibility with Julia v0.6:
-
The syntax for elementwise multiplication, division, and exponentiation has changed. In place of
x .* y
the syntax is now
dot(*)(x, y)
Explicitly calling
broadcast(*, x, y)
also works. Similarly, elementwise division is invoked asdot(/)
, and elementwise exponentiation isdot(^)
. -
The previous syntax (
.*
,./
, and.^
) still works on earlier versions of Julia. -
We also fixed some deprecation warnings.
Support for Complex Variables and Coefficients
This release adds a major new feature: we now support complex variables, functions of complex variables, and complex coefficients. For details, see the documentation.
We also have some small changes: this release
- enables precompilation
- bug fixes