chore: use air formatting

.Rbuildignore

@@ -6,6 +6,7 @@
 ^Meta$
 ^README\.Rmd$
 ^README\.html$
+^[\.]?air\.toml$
 ^\.Renviron$
 ^\.Rproj\.user$
 ^\.ccache$

.pre-commit-config.yaml

@@ -4,10 +4,6 @@ repos:
   - repo: https://github.com/lorenzwalthert/precommit
     rev: v0.4.3.9008
     hooks:
-      - id: style-files
-        args: [--style_pkg=styler.mlr, --style_fun=mlr_style]
-        additional_dependencies:
-          - mlr-org/styler.mlr
       - id: roxygenize
         additional_dependencies:
           - R6

R/PipeOpFDACor.R

@@ -23,7 +23,8 @@
 #' po_cor = po("fda.cor")
 #' task_cor = po_cor$train(list(task))[[1L]]
 #' task_cor
-PipeOpFDACor = R6Class("PipeOpFDACor",
+PipeOpFDACor = R6Class(
+  "PipeOpFDACor",
   inherit = PipeOpTaskPreprocSimple,
   public = list(
     #' @description Initializes a new instance of this Class.

R/PipeOpFDAExtract.R

@@ -49,7 +49,8 @@
 #' )
 #' task_custom = po_custom$train(list(task))[[1L]]
 #' task_custom
-PipeOpFDAExtract = R6Class("PipeOpFDAExtract",
+PipeOpFDAExtract = R6Class(
+  "PipeOpFDAExtract",
   inherit = PipeOpTaskPreprocSimple,
   public = list(
     #' @description Initializes a new instance of this Class.
@@ -63,41 +64,44 @@ PipeOpFDAExtract = R6Class("PipeOpFDAExtract",
         drop = p_lgl(tags = c("train", "predict", "required")),
         left = p_dbl(tags = c("train", "predict", "required")),
         right = p_dbl(tags = c("train", "predict", "required")),
-        features = p_uty(tags = c("train", "predict", "required"), custom_check = crate(function(x) {
-          if (test_character(x)) {
-            return(check_subset(x, choices = c("mean", "median", "min", "max", "slope", "var")))
-          }
-          if (test_list(x)) {
-            res = check_list(x, types = c("character", "function"), any.missing = FALSE, unique = TRUE)
-            if (!isTRUE(res)) {
-              return(res)
+        features = p_uty(
+          tags = c("train", "predict", "required"),
+          custom_check = crate(function(x) {
+            if (test_character(x)) {
+              return(check_subset(x, choices = c("mean", "median", "min", "max", "slope", "var")))
             }
-            nms = names2(x)
-            res = check_names(nms[!is.na(nms)], "unique")
-            if (!isTRUE(res)) {
-              return(res)
-            }
-            for (i in seq_along(x)) {
-              if (is.function(x[[i]])) {
-                res = check_function(x[[i]], args = c("arg", "value"))
-                if (!isTRUE(res)) {
-                  return(res)
-                }
-                res = check_names(nms[i])
-                if (!isTRUE(res)) {
-                  return(res)
-                }
-              } else {
-                res = check_choice(x[[i]], choices = c("mean", "median", "min", "max", "slope", "var"))
-                if (!isTRUE(res)) {
-                  return(res)
+            if (test_list(x)) {
+              res = check_list(x, types = c("character", "function"), any.missing = FALSE, unique = TRUE)
+              if (!isTRUE(res)) {
+                return(res)
+              }
+              nms = names2(x)
+              res = check_names(nms[!is.na(nms)], "unique")
+              if (!isTRUE(res)) {
+                return(res)
+              }
+              for (i in seq_along(x)) {
+                if (is.function(x[[i]])) {
+                  res = check_function(x[[i]], args = c("arg", "value"))
+                  if (!isTRUE(res)) {
+                    return(res)
+                  }
+                  res = check_names(nms[i])
+                  if (!isTRUE(res)) {
+                    return(res)
+                  }
+                } else {
+                  res = check_choice(x[[i]], choices = c("mean", "median", "min", "max", "slope", "var"))
+                  if (!isTRUE(res)) {
+                    return(res)
+                  }
                 }
               }
+              return(TRUE)
             }
-            return(TRUE)
-          }
-          "Features must be a character or list"
-        }))
+            "Features must be a character or list"
+          })
+        )
       )
       param_set$set_values(
         drop = TRUE,
@@ -144,7 +148,8 @@ PipeOpFDAExtract = R6Class("PipeOpFDAExtract",
         if (is.function(feature)) {
           return(feature)
         }
-        switch(feature,
+        switch(
+          feature,
           mean = fmean,
           median = fmedian,
           min = fmin,

R/PipeOpFDAFlatten.R

@@ -21,7 +21,8 @@
 #' task = tsk("fuel")
 #' pop = po("fda.flatten")
 #' task_flat = pop$train(list(task))
-PipeOpFDAFlatten = R6Class("PipeOpFDAFlatten",
+PipeOpFDAFlatten = R6Class(
+  "PipeOpFDAFlatten",
   inherit = PipeOpTaskPreprocSimple,
   public = list(
     #' @description Initializes a new instance of this Class.

R/PipeOpFDAInterpol.R

@@ -45,7 +45,8 @@
 #' pop = po("fda.interpol")
 #' task_interpol = pop$train(list(task))[[1L]]
 #' task_interpol$data()
-PipeOpFDAInterpol = R6Class("PipeOpFDAInterpol",
+PipeOpFDAInterpol = R6Class(
+  "PipeOpFDAInterpol",
   inherit = PipeOpTaskPreprocSimple,
   public = list(
     #' @description Initializes a new instance of this Class.
@@ -56,17 +57,22 @@ PipeOpFDAInterpol = R6Class("PipeOpFDAInterpol",
     #'   otherwise be set during construction. Default `list()`.
     initialize = function(id = "fda.interpol", param_vals = list()) {
       param_set = ps(
-        grid = p_uty(tags = c("train", "predict", "required"), custom_check = crate(function(x) {
-          if (test_string(x)) {
-            return(check_choice(x, choices = c("union", "intersect", "minmax")))
-          }
-          if (test_numeric(x, any.missing = FALSE, min.len = 1L)) {
-            return(TRUE)
-          }
-          "Must be either a string or numeric vector"
-        })),
+        grid = p_uty(
+          tags = c("train", "predict", "required"),
+          custom_check = crate(function(x) {
+            if (test_string(x)) {
+              return(check_choice(x, choices = c("union", "intersect", "minmax")))
+            }
+            if (test_numeric(x, any.missing = FALSE, min.len = 1L)) {
+              return(TRUE)
+            }
+            "Must be either a string or numeric vector"
+          })
+        ),
         method = p_fct(
-          c("linear", "spline", "fill_extend", "locf", "nocb"), default = "linear", tags = c("train", "predict")
+          c("linear", "spline", "fill_extend", "locf", "nocb"),
+          default = "linear",
+          tags = c("train", "predict")
         ),
         left = p_dbl(tags = c("train", "predict")),
         right = p_dbl(tags = c("train", "predict"))

R/PipeOpFDAScaleRange.R

@@ -21,7 +21,8 @@
 #' po_scale = po("fda.scalerange", lower = -1, upper = 1)
 #' task_scale = po_scale$train(list(task))[[1L]]
 #' task_scale$data()
-PipeOpFDAScaleRange = R6Class("PipeOpFDAScaleRange",
+PipeOpFDAScaleRange = R6Class(
+  "PipeOpFDAScaleRange",
   inherit = PipeOpTaskPreproc,
   public = list(
     #' @description Initializes a new instance of this Class.

R/PipeOpFDASmooth.R

@@ -31,7 +31,8 @@
 #' task_smooth = po_smooth$train(list(task))[[1L]]
 #' task_smooth
 #' task_smooth$data(cols = c("NIR", "UVVIS"))
-PipeOpFDASmooth = R6Class("PipeOpFDASmooth",
+PipeOpFDASmooth = R6Class(
+  "PipeOpFDASmooth",
   inherit = PipeOpTaskPreprocSimple,
   public = list(
     #' @description Initializes a new instance of this Class.
@@ -42,9 +43,14 @@ PipeOpFDASmooth = R6Class("PipeOpFDASmooth",
     #'   otherwise be set during construction. Default `list()`.
     initialize = function(id = "fda.smooth", param_vals = list()) {
       param_set = ps(
-        method = p_fct(default = "lowess", c("lowess", "rollmean", "rollmedian", "savgol"), tags = c("train", "predict")), # nolint
+        method = p_fct(
+          default = "lowess",
+          c("lowess", "rollmean", "rollmedian", "savgol"),
+          tags = c("train", "predict")
+        ),
         args = p_uty(
-          tags = c("train", "predict", "required"), custom_check = crate(function(x) check_list(x, names = "unique"))
+          tags = c("train", "predict", "required"),
+          custom_check = crate(function(x) check_list(x, names = "unique"))
         ),
         verbose = p_lgl(tags = c("train", "predict", "required"))
       )
@@ -72,7 +78,8 @@ PipeOpFDASmooth = R6Class("PipeOpFDASmooth",
         return(dt)
       }
       for (j in seq_along(dt)) {
-        set(dt,
+        set(
+          dt,
           j = j,
           value = suppressMessages(invoke(tf::tf_smooth, x = dt[[j]], method = pars$method, .args = pars$args))
         )

R/PipeOpFDATsfeatures.R

@@ -38,7 +38,8 @@
 #' po_tsfeats = po("fda.tsfeats")
 #' task_tsfeats = po_tsfeats$train(list(task))[[1L]]
 #' task_tsfeats$data()
-PipeOpFDATsfeatures = R6Class("PipeOpFDATsfeatures",
+PipeOpFDATsfeatures = R6Class(
+  "PipeOpFDATsfeatures",
   inherit = PipeOpTaskPreprocSimple,
   public = list(
     #' @description Initializes a new instance of this Class.

R/PipeOpFPCA.R

@@ -28,7 +28,8 @@
 #' po_fpca = po("fda.fpca", n_components = 3L)
 #' task_fpca = po_fpca$train(list(task))[[1L]]
 #' task_fpca$data()
-PipeOpFPCA = R6Class("PipeOpFPCA",
+PipeOpFPCA = R6Class(
+  "PipeOpFPCA",
   inherit = PipeOpTaskPreproc,
   public = list(
     #' @description Initializes a new instance of this Class.

R/bibentries.R

@@ -1,7 +1,8 @@
 #' @importFrom utils bibentry
 # nolint start
 bibentries = c(
-  ferraty2003curves = bibentry("article",
+  ferraty2003curves = bibentry(
+    "article",
     title = "Curves discrimination: a nonparametric functional approach",
     author = "Ferraty, Fr\xc3\xa9dric and Vieu, Philippe",
     journal = "Computational Statistics & Data Analysis",
@@ -11,7 +12,8 @@ bibentries = c(
     year = "2003",
     publisher = "Elsevier"
   ),
-  brockhaus2015functional = bibentry("article",
+  brockhaus2015functional = bibentry(
+    "article",
     title = "The functional linear array model",
     author = "Brockhaus, Sarah and Scheipl, Fabian and Hothorn, Torsten and Greven, Sonja",
     journal = "Statistical Modelling",
@@ -21,7 +23,8 @@ bibentries = c(
     year = "2015",
     publisher = "SAGE Publications Sage India: New Delhi, India"
   ),
-  goldsmith2011penalized = bibentry("article",
+  goldsmith2011penalized = bibentry(
+    "article",
     title = "Penalized functional regression",
     author = "Goldsmith, Jeff and Bobb, Jennifer and Crainiceanu, Ciprian M and Caffo, Brian and Reich, Daniel",
     journal = "Journal of Computational and Graphical Statistics",

R/zzz.R

@@ -64,7 +64,7 @@ register_mlr3pipelines = function() {
   register_namespace_callback(pkgname, "mlr3pipelines", register_mlr3pipelines)
 }
 
-.onUnload = function(libPaths) { # nolint
+.onUnload = function(libPaths) {
   walk(names(mlr3fda_tasks), function(nm) mlr_tasks$remove(nm))
   walk(names(mlr3fda_pipeops), function(nm) mlr_pipeops$remove(nm))
   mlr_reflections$task_feature_types =

air.toml

@@ -0,0 +1,2 @@
+[format]
+line-width = 120

inst/bench/bench.R

@@ -5,9 +5,7 @@ library(mlr3pipelines)
 set.seed(1234)
 setDTthreads(threads = 1)
 
-generate_data = function(n_patients = 100,
-  n_weeks = 10,
-  type = c("reg", "irreg")) {
+generate_data = function(n_patients = 100, n_weeks = 10, type = c("reg", "irreg")) {
   type = match.arg(type)
   if (type == "reg") {
     week = 1:n_weeks
@@ -52,27 +50,29 @@ analyse_dplyr = function(patients, window_start, window_end, choice) {
 
 analyse_dt = function(patients, window_start, window_end, choice) {
   if (choice == "one") {
-    patients[between(week, window_start, window_end), .(
-      mean = mean(measurement_value)
-    ), keyby = .(patient_id, measurement_type)]
+    patients[
+      between(week, window_start, window_end),
+      .(
+        mean = mean(measurement_value)
+      ),
+      keyby = .(patient_id, measurement_type)
+    ]
   } else {
-    patients[between(week, window_start, window_end), .(
-      mean = mean(measurement_value),
-      var = var(measurement_value),
-      slope = coef(lm(measurement_value ~ week))[[2]]
-    ), keyby = .(patient_id, measurement_type)]
+    patients[
+      between(week, window_start, window_end),
+      .(
+        mean = mean(measurement_value),
+        var = var(measurement_value),
+        slope = coef(lm(measurement_value ~ week))[[2]]
+      ),
+      keyby = .(patient_id, measurement_type)
+    ]
   }
 }
 
 build_graph = function(left, right, choice) {
   features = switch(choice, one = "mean", three = c("mean", "var", "slope"))
-  po("ffs",
-    features = features,
-    id = "features",
-    drop = FALSE,
-    left = left,
-    right = right
-  )
+  po("ffs", features = features, id = "features", drop = FALSE, left = left, right = right)
 }
 
 analyse_fda = function(graph, task) {
@@ -88,7 +88,9 @@ results_one = bench::press(
     window_end = floor(n_weeks * 0.8)
 
     patients = generate_data(
-      n_patients = n_patients, n_weeks = n_weeks, type = type
+      n_patients = n_patients,
+      n_weeks = n_weeks,
+      type = type
     )
     patients_long = patients |>
       tidyr::pivot_longer(
@@ -129,7 +131,9 @@ results_three = bench::press(
     window_end = floor(n_weeks * 0.8)
 
     patients = generate_data(
-      n_patients = n_patients, n_weeks = n_weeks, type = type
+      n_patients = n_patients,
+      n_weeks = n_weeks,
+      type = type
     )
     patients_long = patients |>
       tidyr::pivot_longer(

tests/testthat/test_PipeOpFDAExtract.R

@@ -50,7 +50,10 @@ test_that("PipeOpFDAExtract works", {
   pop = po("fda.extract", features = c("mean", "median", "min"), drop = TRUE, left = 100, right = 200)
   task_pop = train_pipeop(pop, list(task))[[1L]]
   expected = data.table(
-    y = 1:2, f_mean = rep(NA_real_, 2L), f_median = rep(NA_real_, 2L), f_min = rep(NA_real_, 2L)
+    y = 1:2,
+    f_mean = rep(NA_real_, 2L),
+    f_median = rep(NA_real_, 2L),
+    f_min = rep(NA_real_, 2L)
   )
   expect_identical(task_pop$data(), expected)
 
@@ -184,9 +187,7 @@ test_that("ffind works", {
   expect_identical(ffind(1:10, 5, 15), c(5L, 10L))
 
   expect_identical(
-    ffind(c(-3876, -3798, -3453, -3363, -2974, -2953, -2871, -1917, -1335, -1304, -725, 10),
-      left = -200, right = 0
-    ),
+    ffind(c(-3876, -3798, -3453, -3363, -2974, -2953, -2871, -1917, -1335, -1304, -725, 10), left = -200, right = 0),
     rep(NA_integer_, 2L)
   )
 })