0.18.0

What's Changed

• gp, moe, egobox (Python): Rename predict_derivatives() as predict_gradients() by @relf in #148
• Prepare release 0.18 by @relf in #149

Full Changelog: 0.17.0...0.18.0

0.17.0

What's Changed

• [Breaking changes] gp API renaming by @relf in #145
  • predict_values() is renamed predict()
  • predict_variances() is renamed predict_var()
  • predict_variance_derivatives() is renamed predict_var_derivatives()
    Derivatives predictions (predict_derivatives() and predict_var_derivatives()) are made available in Python.
• Refactor Mixture of Experts by @relf in #146
  Factorize code between full GP and sparse GP implementations
• Prepare release 0.17 by @relf in #147

Full Changelog: 0.16.0...0.17.0

0.16.0

What's Changed

• Add Gpx accessors by @relf in #140
  Some surrogate traits where refactored:
  • Add GpSurrogateExt trait to manage derivatives and sampling
  • Add GpParameterized trait to manage optimized parameters
  • MixtureGpSurrogate is not a FullGpSurrogate any more
  • Add experts retrieval from GpMixture
• Fix LHS maximin bug by @relf in #141

Full Changelog: 0.15.3...0.16.0

0.15.3

What's Changed

• Improve classic, centered and maximin LHS performance by @relf in #138
• Prepare release 0.15.3 by @relf in #139

Full Changelog: 0.15.2...0.15.3

0.15.2

What's Changed

• doe: Improve optimized LHS performance (1.25x speedup) by @relf in #136
• Prepare release 0.15.2 by @relf in #137

Full Changelog: 0.15.1...0.15.2

0.15.1

What's Changed

• Rework (mostly internal) API to avoid awkward &Option by @relf in #134
• Add Python bindings for all LHS flavours by @relf in #135

Full Changelog: 0.15.0...0.15.1

0.15.0

What's Changed

• Implement Sparse gaussian process methods: see tutorial
  • Implement sparse gaussian process methods by @relf in #128
  • Rework SGP API by @relf in #129
  • Rework Python GpMix API by @relf in #130 Breaking Changes
  • Mixture of sparse GP by @relf in #131
  • Improve GP/SGP API by @relf in #132 Breaking Changes
• Re-introduce LHS optim parallel processing by @relf in #123
• Build and release on macos by @relf in #124
• Maintainance by @relf in #125

Full Changelog: 0.14.0...0.15.0

0.14.0

What's Changed

• Fix Egor service in presence of discrete variables by @relf in #119
  A few API breaking changes:
• EgorConfig::xtypes not an option anymore
• EgorSolver::new_with_xtypes() renamed new as new with xlimits is removed, use to_xtypes to convert xlimits
• EgorConfig::no_discrete attribute removed, use EgorConfig::discrete() method
• SurrogateBuilder::new_with_xtypes_rng renamed new_with_xtypes

Full Changelog: 0.13.0...0.14.0

0.13.0

ego: Egor ask-and-tell interface by @relf in #116

Breaking Changes: API refactoring to enable ask-and-tell interface

• Configuration of Egor is factorized out in EgorConfig
• EgorBuilder gets a configure method to tune the configuration
• EgorService structure represents Egor when used as service (ie. ask-and-tell API)
• Python Egor API changes:
  • function under optimization is now pass via minimize(fun, max_iters=...) method
  • new method suggest(xdoe, ydoe) allows to ask for x suggestion and tell current function evaluations
  • new method get_result(xdoe, ydoe) to get the best evaluation (ie the minimum) from given ones

Full Changelog: 0.12.0...0.13.0

0.12.0

What's Changed

• gp now uses cobyla by @relf in #110 , #113
• ego now uses cobyla and slsqp (nlopt is made optional) by @relf in #112
• egobox Python module: Better mixed-integer type declaration by @relf in #115
• Upgrade dependencies by @relf in #114
• Upgrade edition 2021 by @relf in #109
• CI maintainance by @relf in #111
• Bump actions/checkout from 2 to 4 by @dependabot in #107
• Bump actions/setup-python from 2 to 4 by @dependabot in #108

Full Changelog: 0.11.0...0.12.0