search

astamm / nloptr

nloptr provides an R interface to NLopt, a free/open-source library for nonlinear optimization providing a common interface to a number of different optimization routines which can handle nonlinear constraints and lower and upper bounds for the controls.
https://astamm.github.io/nloptr/
Other
103 stars 36 forks source link

Reverse direction of hin/hinjac in wrapper functions and update various examples. #157

Closed aadler closed 2 weeks ago

aadler commented 3 weeks ago

This PR should address issues #148, #152, and #156. It reverses the expected inequality direction of hin/hinjac without changing the default behavior by adding a deprecatedBehavior parameter, whose default is TRUE. The idea being a future (major/breaking) release will change the default behavior to FALSE and a subsequent release will remove the option completly. It also cleans up many, but not all, of examples. It combines all the changes since the last push to CRAN (2.0.3) nto one release (2.1.0) in NEWS.md, which has both a synopsis of the major change and collected bullet points.

aadler commented 3 weeks ago

As this wasn't merged yet, I upgraded all the deprecated messages to warnings and brought the SLSQP test suite in line with the remainders, in that the new behavior is tested first and the old at the end. This way, when we finally remove the old behavior it will be simpler to prune the test suite.

astamm commented 2 weeks ago

Awesome! that answers my various comments in the issues! Thanks again @aadler !