search

luinardi / hypermapper

Black-box Optimizer based on Bayesian Optimization
MIT License
155 stars 26 forks source link

How to find the samples of finding answer #80

Closed peterhchen closed 2 years ago

peterhchen commented 2 years ago

The OpenTurner counts for unique samplesseens and minimum penalty found (3214, 1.05000) (4566, 0.70000) (8453, 0.65000) (10506, 0.55000) (10981, 0.50000) ....

Right now, I used HM_CNT and MINI_DIST to keep track of minimum penalty. Is there a way that hypermapper can print out the similar output? My test case is attached as below/ https://github.com/peterhchen/960_Hypermapper Sincerely Yours, Peter H. Chen

arturluis commented 2 years ago

Hi @peterhchen ,

By default, HyperMapper only explores unique points every iteration, there are no duplicates. So the number of unique samples seen should be the same as the total number of samples seen.

Also, the best point found is printed at the end of execution. HyperMapper prints "Best point found:" followed by the input parameters that lead to the minimum penalty and the minimum penalty value.

Best, Artur Souza

peterhchen commented 2 years ago

Thanks

On Wed, Aug 10, 2022 at 12:34 PM arturluis @.***> wrote:

Hi @peterhchen https://github.com/peterhchen ,

By default, HyperMapper only explores unique points every iteration, there are no duplicates. So the number of unique samples seen should be the same as the total number of samples seen.

Also, the best point found is printed at the end of execution. HyperMapper prints "Best point found:" followed by the input parameters that lead to the minimum penalty and the minimum penalty value.

Best, Artur Souza

— Reply to this email directly, view it on GitHub https://github.com/luinardi/hypermapper/issues/80#issuecomment-1211179618, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADBSBD4JAYZ4O7CA7UHALGLVYP753ANCNFSM55W23VMQ . You are receiving this because you were mentioned.Message ID: @.***>

-- Sincerely Yours, Peter H. Chen