clean up some imports

Author: dajmcdon

DESCRIPTION

@@ -18,6 +18,7 @@ Imports:
     cli,
     dplyr,
     epiprocess,
+    generics,
     glue,
     hardhat (>= 1.0.0.9000),
     magrittr,

NAMESPACE

@@ -45,7 +45,8 @@ export(smooth_arx_forecaster)
 export(step_epi_ahead)
 export(step_epi_lag)
 import(recipes)
-importFrom(broom,augment)
+importFrom(generics,augment)
+importFrom(generics,fit)
 importFrom(magrittr,"%>%")
 importFrom(rlang,"!!")
 importFrom(rlang,":=")
@@ -60,4 +61,5 @@ importFrom(stats,predict)
 importFrom(stats,quantile)
 importFrom(stats,residuals)
 importFrom(stats,setNames)
+importFrom(tibble,as_tibble)
 importFrom(tibble,tibble)

R/epi_recipe.R

@@ -417,6 +417,7 @@ kill_levels <- function(x, keys) {
   x
 }
 
+#' @importFrom tibble as_tibble
 #' @export
 as_tibble.epi_df <- function(x, ...) {
   # so that downstream calls to as_tibble don't clobber our metadata

R/epi_workflow.R

@@ -13,6 +13,9 @@
 #' @return A new `epi_workflow` object.
 #' @seealso workflows::workflow
 #' @importFrom rlang is_null
+#' @importFrom stats predict
+#' @importFrom generics fit
+#' @importFrom generics augment
 #' @export
 #' @examples
 #' library(dplyr)
@@ -168,7 +171,6 @@ grab_forged_keys <- function(forged, mold, new_data) {
 #' @param ... Arguments passed on to the predict method.
 #'
 #' @return new_data with additional columns containing the predicted values
-#' @importFrom broom augment
 #' @export
 augment.epi_workflow <- function (x, new_data, ...) {
   predictions <- predict(x, new_data, ...)

R/layer_predict.R

@@ -1,21 +1,26 @@
 layer_predict <-
-  function(frosting, id = rand_id("predict_default")) {
+  function(frosting, type = NULL, opts = list(), ..., id = rand_id("predict_default")) {
     add_layer(
       frosting,
       layer_predict_new(
+        type = type,
+        opts = opts,
+        dots_list = rlang::list2(...), # can't figure how to use this
         id = id
       )
     )
   }
 
 
-layer_predict_new <- function(id) {
-  layer("predict", id = id)
+layer_predict_new <- function(type, opts, dots_list, id) {
+  layer("predict", type = type, opts = opts, dots_list = dots_list, id = id)
 }
 
 #' @export
 slather.layer_predict <- function(object, components, the_fit) {
-  components$predictions <- predict(the_fit, components$forged$predictors)
+
+  components$predictions <- predict(the_fit, components$forged$predictors,
+                                    type = object$type, opts = object$opts)
   components$predictions <- dplyr::bind_cols(components$keys, components$predictions)
   components
 }

tests/testthat/test-frosting.R

@@ -24,11 +24,13 @@ test_that("prediction works without any postprocessor", {
     step_epi_ahead(death_rate, ahead = 7) %>%
     step_naomit(all_predictors()) %>%
     step_naomit(all_outcomes(), skip = TRUE)
-  wf <- epi_workflow(r, linear_reg()) %>% fit(jhu)
-  latest <- jhu %>% filter(time_value >= max(time_value) - 14)
+  wf <- epi_workflow(r, parsnip::linear_reg()) %>% fit(jhu)
+  latest <- jhu %>%
+    dplyr::filter(time_value >= max(time_value) - 14)
 
   expect_silent(predict(wf, latest))
-  p <- predict(wf, latest) %>% dplyr::filter(!is.na(.pred))
+  p <- predict(wf, latest) %>%
+    dplyr::filter(!is.na(.pred))
   expect_equal(nrow(p), 3)
   expect_s3_class(p, "epi_df")
   expect_equal(tail(p$time_value, 1), as.Date("2021-12-31"))

tests/testthat/test-layer_predict.R

@@ -6,8 +6,9 @@ test_that("predict layer works alone", {
     step_epi_ahead(death_rate, ahead = 7) %>%
     step_naomit(all_predictors()) %>%
     step_naomit(all_outcomes(), skip = TRUE)
-  wf <- epi_workflow(r, linear_reg()) %>% fit(jhu)
-  latest <- jhu %>% filter(time_value >= max(time_value) - 14)
+  wf <- epi_workflow(r, parsnip::linear_reg()) %>% fit(jhu)
+  latest <- jhu %>%
+    dplyr::filter(time_value >= max(time_value) - 14)
 
   f <- frosting() %>% layer_predict()
   wf <- wf %>% add_frosting(f)