Allow package version designation for CRAN packages w/ estimator

NAMESPACE

@@ -12,6 +12,7 @@ export(complete_run)
 export(container_registry)
 export(convert_to_dataset_with_csv_files)
 export(convert_to_dataset_with_parquet_files)
+export(cran_package)
 export(create_aks_compute)
 export(create_aml_compute)
 export(create_child_run)

R/environment.R

@@ -256,29 +256,71 @@ generate_docker_file <- function(custom_docker_image = NULL,
 
   if (!is.null(cran_packages)) {
     for (package in cran_packages) {
+
+      package$repo <- sprintf("c(\'%s\')", package$repo)
+
       base_dockerfile <- paste0(
-          base_dockerfile,
-          sprintf("RUN R -e \"install.packages(\'%s\', ", package),
-          "repos = \'https://cloud.r-project.org/\')\"\n")
+        base_dockerfile,
+        paste0("RUN R -e \"remotes::install_version",
+               sprintf("(\'%s\'", package$name),
+               sprintf(", version = \'%s\'", package$version),
+               sprintf(", repos = %s", package$repo),
+               ")\"\n"
+        )
+      )
     }
   }
 
-  if (!is.null(github_packages)) {
-    for (package in github_packages) {
-      base_dockerfile <- paste0(
-          base_dockerfile,
-          sprintf("RUN R -e \"remotes::install_github(\'%s\')\"\n", package))
-    }
+
+if (!is.null(github_packages)) {
+  for (package in github_packages) {
+    base_dockerfile <- paste0(
+      base_dockerfile,
+      sprintf("RUN R -e \"remotes::install_github(\'%s\')\"\n", package))
   }
+}
 
-  if (!is.null(custom_url_packages)) {
-    for (package in custom_url_packages) {
-      base_dockerfile <- paste0(
-          base_dockerfile,
-          sprintf("RUN R -e \"install.packages(\'%s\', repos = NULL)\"\n",
-                  package))
-    }
+if (!is.null(custom_url_packages)) {
+  for (package in custom_url_packages) {
+    base_dockerfile <- paste0(
+      base_dockerfile,
+      sprintf("RUN R -e \"install.packages(\'%s\', repos = NULL)\"\n",
+              package))
   }
+}
+
+invisible(base_dockerfile)
+}
+
+#' Specifies a CRAN package to install in environment
+#'
+#' @description
+#' Specifies a CRAN package to install in run environment
+#'
+#' @param name The package name
+#' @param version A string of the package version. If not provided, version
+#' will default to latest
+#' @param repo The base URL of the repository to use, e.g., the URL of a
+#' CRAN mirror. If not provided, the package will be pulled from
+#' "https://cloud.r-project.org".
+#' @return A named list containing the package specifications
+#' @export
+#' @section Examples:
+#' ```
+#' pkg1 <- cran_package("ggplot2", version = "3.3.0")
+#' pkg2 <- cran_package("stringr")
+#' pkg3 <- cran_package("ggplot2", version = "0.9.1",
+#'                      repo = "http://cran.us.r-project.org")
+#'
+#' est <- estimator(source_directory = ".",
+#'                  entry_script = "train.R",
+#'                  cran_packages = list(pkg1, pkg2, pkg3),
+#'                  compute_target = compute_target)
+#' ```
+#' @seealso [r_environment()], [estimator()]
+#' @md
+cran_package <- function(name, version = NULL, repo = "https://cloud.r-project.org") {
+  cran_package <- list(name = name, version = version, repo = repo)
 
-  invisible(base_dockerfile)
+  return(cran_package)
 }

R/estimator.R

@@ -71,11 +71,20 @@
 #' objects to use as input.
 #' @return The `Estimator` object.
 #' @export
+#' @section Examples:
+#' ```
+#' pkg1 <- cran_package("ggplot2", version = "3.3.0")
+#' pkg2 <- cran_package("stringr")
+#'
+#' est <- estimator(source_directory = ".",
+#'                  entry_script = "train.R",
+#'                  cran_packages = list(pkg1, pkg2),
+#'                  compute_target = compute_target)
+#' ```
 #' @seealso
-#' \code{\link{r_environment}}
-#' \code{\link{container_registry}}
-#' \code{\link{submit_experiment}}
-#' \code{\link{dataset_consumption_config}}
+#' [r_environment()], [container_registry()], [submit_experiment()],
+#' [dataset_consumption_config()], [cran_package()]
+#'
 #' @md
 estimator <- function(source_directory,
                       compute_target = NULL,

_pkgdown.yml

@@ -158,6 +158,7 @@ reference:
   - '`log_row_to_run`'
   - '`log_table_to_run`'
   - '`view_run_details`'
+  - '`cran_package`'
 - title: Hyperparameter tuning
   desc: Functions for configuring and managing hyperparameter tuning (HyperDrive) experiments. Azure ML's HyperDrive functionality enables you to automate hyperparameter tuning of your machine learning models. For example, you can define the parameter search space as discrete or continuous, and a sampling method over the search space as random, grid, or Bayesian. Also, you can specify a primary metric to optimize in the hyperparameter tuning experiment, and whether to minimize or maximize that metric. You can also define early termination policies in which poorly performing experiment runs are canceled and new ones started.
   contents:

tests/testthat/test_environment.R

@@ -12,7 +12,7 @@ test_that("create environment and check parameters", {
   expect_equal(env$version, "1")
   expect_equal(env$docker$enabled, TRUE)
   expect_equal(env$docker$base_image, NULL)
-
+  
   # use custom docker image
   custom_docker_image_name = "temp_image"
   env <- r_environment(env_name, custom_docker_image = custom_docker_image_name)
@@ -25,15 +25,15 @@ test_that("create environment and check parameters", {
 test_that("create, register, and get environment", {
   skip_if_no_subscription()
   ws <- existing_ws
-
+  
   env_name <- "testenv"
-
+  
   # Create environment
   env <- r_environment(env_name, version = "1")
-
+  
   # Register environment
   register_environment(env, ws)
-
+  
   # Get environment
   environ <- get_environment(ws, env_name, "1")
   expect_equal(env$name, environ$name)
@@ -49,30 +49,37 @@ test_that("create dockerfile", {
                                   "RUN R -e \"remotes::install_cran('azuremlsdk'",
                                   ", repos = 'https://cloud.r-project.org/', ",
                                   "upgrade = FALSE)\"\n"))
-
+  
   # cran packages
+  cran_pkg_1 <- cran_package("ggplot2")
+  cran_pkg_2 <- cran_package("stringr", "0.0.0", "https://some-other-repo.com/")
+  cran_pkg_3 <- cran_package('shiny', "1.14")
   dockerfile <- generate_docker_file(custom_docker_image = "ubuntu-18.04",
-                                     cran_packages = c("ggplot2"),
+                                     cran_packages = list(cran_pkg_1,
+                                                          cran_pkg_2,
+                                                          cran_pkg_3),
                                      install_system_packages = FALSE)
-  expect_equal(dockerfile, paste0("FROM ubuntu-18.04\nRUN R -e \"install.",
-                                  "packages('ggplot2', repos = \'https://",
-                                  "cloud.r-project.org/\')\"\n"))
-
+  expected_dockerfile <- paste0("FROM ubuntu-18.04\n",
+                                "RUN R -e \"remotes::install_version('ggplot2', repos = c('https://cloud.r-project.org'))\"\n",
+                                "RUN R -e \"remotes::install_version('stringr', version = '0.0.0', repos = c('https://some-other-repo.com/'))\"\n",
+                                "RUN R -e \"remotes::install_version('shiny', version = '1.14', repos = c('https://cloud.r-project.org'))\"\n")
+  expect_equal(dockerfile, expected_dockerfile)
+
   # github packages
   dockerfile <- generate_docker_file(github_packages = c(
-                                       "https://github/user/repo1", 
-                                       "https://github/user/repo2"),
-                                     install_system_packages = FALSE)
+    "https://github/user/repo1", 
+    "https://github/user/repo2"),
+    install_system_packages = FALSE)
   expected_dockerfile <- paste0(
     "RUN R -e \"remotes::install_github(\'https://github/user/repo1\')\"\n",
     "RUN R -e \"remotes::install_github(\'https://github/user/repo2\')\"\n")
   expect_equal(dockerfile, expected_dockerfile)
 
   # custom url
   dockerfile <- generate_docker_file(custom_url_packages = c(
-                                       "https://url/pak1.tar", 
-                                       "https://url/pak2.tar"),
-                                     install_system_packages = FALSE)
+    "https://url/pak1.tar", 
+    "https://url/pak2.tar"),
+    install_system_packages = FALSE)
   expected_dockerfile <- paste0(
     "RUN R -e \"install.packages(\'https://url/pak1.tar\', repos = NULL)\"\n",
     "RUN R -e \"install.packages(\'https://url/pak2.tar\', repos = NULL)\"\n")

tests/testthat/test_estimator.R

@@ -2,10 +2,11 @@ context("estimator")
 
 test_that("create estimator", {
   skip_if_no_azureml()
+
   est <- estimator(".",
                    compute_target = "local",
                    script_params = list("param1" = 1),
-                   cran_packages = c("ggplot2"),
+                   cran_packages = list(cran_package("ggplot2")),
                    use_gpu = TRUE,
                    environment_variables = list("var1" = "val1"))
 

vignettes/hyperparameter-tune-with-keras/hyperparameter-tune-with-keras.Rmd

@@ -119,7 +119,7 @@ To create the estimator, define the following:
 est <- estimator(source_directory = ".",
                  entry_script = "cifar10_cnn.R",
                  compute_target = compute_target,
-                 cran_packages = c("keras"),
+                 cran_packages = list(cran_package("keras")),
                  use_gpu = TRUE)
 ```
 

vignettes/train-with-tensorflow/train-with-tensorflow.Rmd

@@ -103,7 +103,7 @@ To create the estimator, define the following:
 est <- estimator(source_directory = ".",
                  entry_script = "tf_mnist.R",
                  compute_target = compute_target,
-                 cran_packages = c("tensorflow"),
+                 cran_packages = list(cran_package("tensorflow")),
                  use_gpu = TRUE)
 ```