chore: sms_ego --> smsego; update DESCRIPTION, add references, add cphs, minimal doc updates

Author: sumny

DESCRIPTION

@@ -1,6 +1,6 @@
 Type: Package
 Package: mlr3mbo
-Title: Flexible Bayesian Optimization in R
+Title: Flexible Bayesian Optimization
 Version: 0.1.1
 Authors@R: c(
     person("Lennart", "Schneider", , "[email protected]", role = c("cre", "aut"),
@@ -17,15 +17,24 @@ Authors@R: c(
            comment = c(ORCID = "0000-0001-8867-762X")),
     person("Martin", "Binder", , "[email protected]", role = "aut"),
     person("Sebastian", "Fischer", , "[email protected]", role = "aut",
-           comment = c(ORCID = "0000-0002-9609-3197")))
-Description: A modern and flexible approach to Bayesian Optimization /
-    Model Based Optimization in R building on the 'bbotk' package. 'mlr3mbo' is
-    a toolbox providing both ready-to-use optimization algorithms as well
-    as their fundamental building blocks. Single- and multi-objective
-    optimization is supported as well as mixed continuous, categorical and
-    conditional search spaces. Moreover, using 'mlr3mbo' for hyperparameter
-    optimization of machine learning models within the 'mlr3' ecosystem is
-    straightforward via 'mlr3tuning'.
+           comment = c(ORCID = "0000-0002-9609-3197")),
+    person("Michael H.", "Buselli", role = "cph"),
+    person("Wessel", "Dankers", role = "cph"),
+    person("Carlos", "Fonseca", role = "cph"),
+    person("Manuel", "Lopez-Ibanez", role = "cph"),
+    person("Luis", "Paquete", role = "cph"))
+Description: A modern and flexible approach to Bayesian Optimization / Model
+    Based Optimization building on the 'bbotk' package. 'mlr3mbo' is a toolbox
+    providing both ready-to-use optimization algorithms as well as their fundamental
+    building blocks allowing for straightforward implementation of custom
+    algorithms. Single- and multi-objective optimization is supported as well as
+    mixed continuous, categorical and conditional search spaces. Moreover, using
+    'mlr3mbo' for hyperparameter optimization of machine learning models within the
+    'mlr3' ecosystem is straightforward via 'mlr3tuning'. Examples of ready-to-use
+    optimization algorithms include Efficient Global Optimization by Jones et al.
+    (1998) <doi:10.1023/A:1008306431147>, ParEGO by Knowles (2006)
+    <doi:10.1109/TEVC.2005.851274> and SMS-EGO by Ponweiser et al. (2008)
+    <doi:10.1007/978-3-540-87700-4_78>.
 License: LGPL-3
 URL: https://mlr3mbo.mlr-org.com, https://github.com/mlr-org/mlr3mbo
 BugReports: https://github.com/mlr-org/mlr3mbo/issues

R/AcqFunctionSmsEgo.R

@@ -1,7 +1,7 @@
 #' @title Acquisition Function SMS-EGO
 #'
 #' @include AcqFunction.R
-#' @name mlr_acqfunctions_sms_ego
+#' @name mlr_acqfunctions_smsego
 #'
 #' @description
 #' S-Metric Selection Evolutionary Multi-Objective Optimization Algorithm Acquisition Function.
@@ -54,7 +54,7 @@
 #'
 #'   surrogate = srlrnc(list(learner, learner$clone(deep = TRUE)), archive = instance$archive)
 #'
-#'   acq_function = acqf("sms_ego", surrogate = surrogate)
+#'   acq_function = acqf("smsego", surrogate = surrogate)
 #'
 #'   acq_function$surrogate$update()
 #'   acq_function$progress = 5 - 4 # n_evals = 5 and 4 points already evaluated
@@ -101,7 +101,7 @@ AcqFunctionSmsEgo = R6Class("AcqFunctionSmsEgo",
       constants$values$lambda = lambda
       constants$values$epsilon = epsilon
 
-      super$initialize("acq_sms_ego", constants = constants, surrogate = surrogate, direction = "minimize", label = "SMS-EGO", man = "mlr3mbo::mlr_acqfunctions_sms_ego")  # indeed, we minimize, see comments below about C code
+      super$initialize("acq_smsego", constants = constants, surrogate = surrogate, direction = "minimize", label = "SMS-EGO", man = "mlr3mbo::mlr_acqfunctions_smsego")  # indeed, we minimize, see comments below about C code
     },
 
     #' @description
@@ -155,10 +155,10 @@ AcqFunctionSmsEgo = R6Class("AcqFunctionSmsEgo",
       # allocate memory for adding points to front for HV calculation in C
       front2 = t(rbind(self$ys_front, 0))
       sms = .Call("c_sms_indicator", PACKAGE = "mlr3mbo", cbs, self$ys_front, front2, self$epsilon, self$ref_point)  # note that the negative indicator is returned from C
-      data.table(acq_sms_ego = sms)
+      data.table(acq_smsego = sms)
     }
   )
 )
 
-mlr_acqfunctions$add("sms_ego", AcqFunctionSmsEgo)
+mlr_acqfunctions$add("smsego", AcqFunctionSmsEgo)
 

R/OptimizerMbo.R

@@ -5,7 +5,7 @@
 #' @description
 #' `OptimizerMbo` class that implements Model Based Optimization (MBO).
 #' The implementation follows a modular layout relying on a [loop_function] determining the MBO flavor to be used, e.g.,
-#' [bayesopt_ego] for sequential single-objective Bayesian Optimization, a [Surrogate], an [AcqFunction], e.g., [AcqFunctionEI] for
+#' [bayesopt_ego] for sequential single-objective Bayesian Optimization, a [Surrogate], an [AcqFunction], e.g., [mlr_acqfunctions_ei] for
 #' Expected Improvement and an [AcqOptimizer].
 #'
 #' MBO algorithms are iterative optimization algorithms that make use of a continuously updated surrogate model built for the objective function.
@@ -23,7 +23,7 @@
 #'     Whether this point was already evaluated. Depends on the `skip_already_evaluated` parameter of the [AcqOptimizer].
 #' @export
 #' @examples
-#' \dontrun{
+#' \donttest{
 #' if (requireNamespace("mlr3learners") &
 #'     requireNamespace("DiceKriging") &
 #'     requireNamespace("rgenoud")) {

R/TunerMbo.R

@@ -9,7 +9,7 @@
 #'
 #' @export
 #' @examples
-#' \dontrun{
+#' \donttest{
 #' if (requireNamespace("mlr3learners") &
 #'     requireNamespace("DiceKriging") &
 #'     requireNamespace("rgenoud")) {

R/bayesopt_ego.R

@@ -4,7 +4,7 @@
 #' @name mlr_loop_functions_ego
 #'
 #' @description
-#' MBO loop function for sequential single-objective Bayesian Optimization.
+#' Loop function for sequential single-objective Bayesian Optimization.
 #' Normally used inside an [OptimizerMbo].
 #'
 #' In each iteration after the initial design, the surrogate and acquisition function are updated and the next candidate
@@ -46,7 +46,7 @@
 #' @family Loop Function
 #' @export
 #' @examples
-#' \dontrun{
+#' \donttest{
 #' if (requireNamespace("mlr3learners") &
 #'     requireNamespace("DiceKriging") &
 #'     requireNamespace("rgenoud")) {

R/bayesopt_mpcl.R

@@ -4,7 +4,7 @@
 #' @name mlr_loop_functions_mpcl
 #'
 #' @description
-#' MBO loop function for single-objective Bayesian Optimization via multipoint constant liar.
+#' Loop function for single-objective Bayesian Optimization via multipoint constant liar.
 #' Normally used inside an [OptimizerMbo].
 #'
 #' In each iteration after the initial design, the surrogate and acquisition function are updated.
@@ -56,7 +56,7 @@
 #' @family Loop Function
 #' @export
 #' @examples
-#' \dontrun{
+#' \donttest{
 #' if (requireNamespace("mlr3learners") &
 #'     requireNamespace("DiceKriging") &
 #'     requireNamespace("rgenoud")) {

R/bayesopt_parego.R

@@ -4,7 +4,7 @@
 #' @name mlr_loop_functions_parego
 #'
 #' @description
-#' MBO loop function for multi-objective Bayesian Optimization via ParEGO.
+#' Loop function for multi-objective Bayesian Optimization via ParEGO.
 #' Normally used inside an [OptimizerMbo].
 #'
 #' In each iteration after the initial design, the observed objective function values are normalized and `q` candidates are
@@ -59,7 +59,7 @@
 #' @family Loop Function
 #' @export
 #' @examples
-#' \dontrun{
+#' \donttest{
 #' if (requireNamespace("mlr3learners") &
 #'     requireNamespace("DiceKriging") &
 #'     requireNamespace("rgenoud")) {

R/bayesopt_smsego.R

@@ -4,11 +4,11 @@
 #' @name mlr_loop_functions_smsego
 #'
 #' @description
-#' MBO loop function for sequential multi-objective Bayesian Optimization via SmsEGO.
+#' Loop function for sequential multi-objective Bayesian Optimization via SMS-EGO.
 #' Normally used inside an [OptimizerMbo].
 #'
-#' In each iteration after the initial design, the surrogate and acquisition function ([AcqFunctionSmsEgo]) are updated and the next candidate
-#' is chosen based on optimizing the acquisition function.
+#' In each iteration after the initial design, the surrogate and acquisition function ([mlr_acqfunctions_smsego]) are
+#' updated and the next candidate is chosen based on optimizing the acquisition function.
 #'
 #' @param instance ([bbotk::OptimInstanceMultiCrit])\cr
 #'   The [bbotk::OptimInstanceMultiCrit] to be optimized.
@@ -19,8 +19,8 @@
 #'   Points are drawn uniformly at random.
 #' @param surrogate ([SurrogateLearnerCollection])\cr
 #'   [SurrogateLearnerCollection] to be used as a surrogate.
-#' @param acq_function ([AcqFunctionSmsEgo])\cr
-#'   [AcqFunctionSmsEgo] to be used as acquisition function.
+#' @param acq_function ([mlr_acqfunctions_smsego])\cr
+#'   [mlr_acqfunctions_smsego] to be used as acquisition function.
 #' @param acq_optimizer ([AcqOptimizer])\cr
 #'   [AcqOptimizer] to be used as acquisition function optimizer.
 #' @param random_interleave_iter (`integer(1)`)\cr
@@ -34,7 +34,7 @@
 #' * The `acq_function$surrogate`, even if already populated, will always be overwritten by the `surrogate`.
 #' * The `acq_optimizer$acq_function`, even if already populated, will always be overwritten by `acq_function`.
 #' * The `surrogate$archive`, even if already populated, will always be overwritten by the [bbotk::Archive] of the [bbotk::OptimInstanceMultiCrit].
-#' * Due to the iterative computation of the epsilon within the [AcqFunctionSmsEgo], requires the [bbotk::Terminator] of
+#' * Due to the iterative computation of the epsilon within the [mlr_acqfunctions_smsego], requires the [bbotk::Terminator] of
 #'   the [bbotk::OptimInstanceMultiCrit] to be a [bbotk::TerminatorEvals].
 #'
 #' @return invisible(instance)\cr
@@ -47,7 +47,7 @@
 #' @family Loop Function
 #' @export
 #' @examples
-#' \dontrun{
+#' \donttest{
 #' if (requireNamespace("mlr3learners") &
 #'     requireNamespace("DiceKriging") &
 #'     requireNamespace("rgenoud")) {
@@ -69,7 +69,7 @@
 #'
 #'   surrogate = default_surrogate(instance)
 #'
-#'   acq_function = acqf("sms_ego")
+#'   acq_function = acqf("smsego")
 #'
 #'   acq_optimizer = acqo(
 #'     optimizer = opt("random_search"),
@@ -146,7 +146,7 @@ bayesopt_smsego = function(
 
 class(bayesopt_smsego) = "loop_function"
 attr(bayesopt_smsego, "id") = "bayesopt_smsego"
-attr(bayesopt_smsego, "label") = "SmsEGO"
+attr(bayesopt_smsego, "label") = "SMS-EGO"
 attr(bayesopt_smsego, "instance") = "multi-crit"
 attr(bayesopt_smsego, "man") = "mlr3mbo::mlr_loop_functions_smsego"
 

R/loop_function.R

@@ -4,7 +4,7 @@
 #' @name loop_function
 #'
 #' @description
-#' Loop functions determine the behavior of the BO algorithm on a global level.
+#' Loop functions determine the behavior of the Bayesian Optimization algorithm on a global level.
 #' For an overview of readily available loop functions, see `as.data.table(mlr_loop_functions)`.
 #'
 #' In general, a loop function is simply a decorated member of the S3 class `loop_function`.

R/mbo_defaults.R

@@ -16,7 +16,7 @@ NULL
 #' @title Default Loop Function
 #'
 #' @description
-#' Chooses a default [loop_function], i.e. the MBO flavor to be used for optimization.
+#' Chooses a default [loop_function], i.e. the Bayesian Optimization flavor to be used for optimization.
 #' For single-objective optimization, defaults to [bayesopt_ego].
 #' For multi-objective optimization, defaults to [bayesopt_smsego].
 #'
@@ -155,6 +155,9 @@ default_surrogate = function(instance, learner = NULL, n_learner = NULL) {
 #'
 #' @description
 #' Chooses a default acquisition function, i.e. the criterion used to propose future points.
+#' For single-objective optimization, defaults to [mlr_acqfunctions_ei].
+#' For multi-objective optimization, defaults to [mlr_acqfunctions_smsego].
+#'
 #' @param instance ([bbotk::OptimInstance]).
 #' @return [AcqFunction]
 #' @family mbo_defaults
@@ -172,14 +175,14 @@ default_acqfun = function(instance) {
 #'
 #' @description
 #' Chooses a default acquisition function optimizer.
-#' Defaults to wrapping [bbotk::OptimizerRandomSearch] allowing 10000 function evaluations.
+#' Defaults to wrapping [bbotk::OptimizerRandomSearch] allowing 10000 function evaluations (with a batch size of 1000) via a [bbotk::TerminatorEvals].
 #'
 #' @param acq_function ([AcqFunction]).
 #' @return [AcqOptimizer]
 #' @family mbo_defaults
 #' @export
 default_acqopt = function(acq_function) {
   assert_r6(acq_function, classes = "AcqFunction")
-  AcqOptimizer$new(optimizer = opt("random_search", batch_size = 10000L), terminator = trm("evals", n_evals = 10000L))  # FIXME: what do we use
+  AcqOptimizer$new(optimizer = opt("random_search", batch_size = 1000L), terminator = trm("evals", n_evals = 10000L))  # FIXME: what do we use
 }