Initial implementation of Constrained GlobalOptimisation #33
Conversation
| @@ -141,17 +138,30 @@ function minimise_icp(f::Function, x::IntervalBox{N, T}; reltol=1e-3, abstol=1e- | |||
| return lb..global_minimum, arg_minima | |||
| end | |||
|
|
|||
| function minimise_icp_constrained(f::Function, x::IntervalBox{N, T}, constraints::Vector{Constraint{T}} = Vector{Constraint{T}}(); reltol=1e-3, abstol=1e-3) where {N, T<:Real} | |||
|
|
|||
| struct ConstraintCond{T} | |||
There was a problem hiding this comment.
Their might be more than one constraint condition thats why we need to built vector of such conditions.
|
|
||
| Q = binary_minheap(IntervalBoxMinimum{N, T}) | ||
|
|
||
| global_minimum = ∞ | ||
|
|
||
| for t in constraints | ||
| x = icp(t.f, x, t.c) | ||
| C = Contractor(t.f, var) |
There was a problem hiding this comment.
It is expensive to create a Contractor, so this should be done only once for each constraint.
There was a problem hiding this comment.
Oh right. We should built contractor for each constraint condition only once.
| @@ -176,10 +186,13 @@ function minimise_icp_constrained(f::Function, x::IntervalBox{N, T}, constraints | |||
| global_minimum = current_minimum | |||
There was a problem hiding this comment.
In general this logic is not correct, since there may be no feasible points in p.
There was a problem hiding this comment.
Yes right..!! this required more work need to be done on this case.
|
Note that if you took code from another commit, you should have built your commit on top of that commit to keep a record of the previous author's contributions. |
|
Yes I'll remember from next time. |
yashvardhan747
changed the title
No description provided.