[R-package] [docs] add intro vignette

.ci/lint_r_code.R

@@ -8,7 +8,7 @@ SOURCE_DIR <- args[[1L]]
 
 FILES_TO_LINT <- list.files(
     path = SOURCE_DIR
-    , pattern = "\\.r$"
+    , pattern = "\\.r$|\\.rmd$"
     , all.files = TRUE
     , ignore.case = TRUE
     , full.names = TRUE

.ci/test_r_package.sh

@@ -92,13 +92,13 @@ if [[ $OS_NAME == "macos" ]]; then
     fi
 fi
 
-# Manually install Depends and Imports libraries + 'testthat'
+# Manually install Depends and Imports libraries + 'knitr', 'rmarkdown', 'testthat'
 # to avoid a CI-time dependency on devtools (for devtools::install_deps())
 # NOTE: testthat is not required when running rchk
 if [[ "${TASK}" == "r-rchk" ]]; then
-    packages="c('data.table', 'jsonlite', 'Matrix', 'R6')"
+    packages="c('data.table', 'jsonlite', 'knitr', 'Matrix', 'R6', 'rmarkdown')"
 else
-    packages="c('data.table', 'jsonlite', 'Matrix', 'R6', 'testthat')"
+    packages="c('data.table', 'jsonlite', 'knitr', 'Matrix', 'R6', 'rmarkdown', 'testthat')"
 fi
 compile_from_source="both"
 if [[ $OS_NAME == "macos" ]]; then

.ci/test_r_package_solaris.sh

@@ -5,7 +5,9 @@ apt-get install --no-install-recommends -y \
   libxml2-dev \
   libssl-dev
 
-Rscript -e "install.packages('rhub', dependencies = c('Depends', 'Imports', 'LinkingTo'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())" || exit -1
+# installation of dependencies needs to happen before building the package,
+# since `R CMD build` needs to install the package to build vignettes
+Rscript -e "install.packages(c('R6', 'data.table', 'jsonlite', 'knitr', 'Matrix', 'rmarkdown', 'rhub', 'testthat'), dependencies = c('Depends', 'Imports', 'LinkingTo'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())" || exit -1
 
 sh build-cran-package.sh || exit -1
 

.ci/test_r_package_valgrind.sh

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-RDscriptvalgrind -e "install.packages(c('R6', 'data.table', 'jsonlite', 'Matrix', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())" || exit -1
+RDscriptvalgrind -e "install.packages(c('R6', 'data.table', 'jsonlite', 'knitr', 'Matrix', 'rmarkdown', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())" || exit -1
 sh build-cran-package.sh \
   --r-executable=RDvalgrind \
   || exit -1

.ci/test_r_package_windows.ps1

@@ -122,7 +122,7 @@ Start-Process -FilePath Rtools.exe -NoNewWindow -Wait -ArgumentList "/VERYSILENT
 Write-Output "Done installing Rtools"
 
 Write-Output "Installing dependencies"
-$packages = "c('data.table', 'jsonlite', 'Matrix', 'processx', 'R6', 'testthat'), dependencies = c('Imports', 'Depends', 'LinkingTo')"
+$packages = "c('data.table', 'jsonlite', 'knitr', 'Matrix', 'processx', 'R6', 'rmarkdown', 'testthat'), dependencies = c('Imports', 'Depends', 'LinkingTo')"
 Run-R-Code-Redirect-Stderr "options(install.packages.check.source = 'no'); install.packages($packages, repos = '$env:CRAN_MIRROR', type = 'binary', lib = '$env:R_LIB_PATH', Ncpus = parallel::detectCores())" ; Check-Output $?
 
 # MiKTeX and pandoc can be skipped on non-MinGW builds, since we don't
@@ -165,7 +165,16 @@ if ($env:COMPILER -ne "MSVC") {
     }
     Run-R-Code-Redirect-Stderr "commandArgs <- function(...){$env:BUILD_R_FLAGS}; source('build_r.R')"; Check-Output $?
   } elseif ($env:R_BUILD_TYPE -eq "cran") {
+    # NOTE: gzip and tar are needed to create a CRAN package on Windows, but
+    # some flavors of tar.exe can fail in some settings on Windows.
+    # Putting the msys64 utilities at the beginning of PATH temporarily to be
+    # sure they're used for that purpose.
+    # $env:PATH = "C:\msys64\usr\bin;" + $env:PATH
+    if ($env:R_MAJOR_VERSION -eq "3") {
+        $env:PATH = "C:\msys64\usr\bin;" + $env:PATH
+    }
     Run-R-Code-Redirect-Stderr "result <- processx::run(command = 'sh', args = 'build-cran-package.sh', echo = TRUE, windows_verbatim_args = FALSE, error_on_status = TRUE)" ; Check-Output $?
+    Remove-From-Path ".*msys64.*"
     # Test CRAN source .tar.gz in a directory that is not this repo or below it.
     # When people install.packages('lightgbm'), they won't have the LightGBM
     # git repo around. This is to protect against the use of relative paths

.github/workflows/r_package.yml

@@ -194,7 +194,7 @@ jobs:
       - name: Install packages
         shell: bash
         run: |
-          RDscript${{ matrix.r_customization }} -e "install.packages(c('R6', 'data.table', 'jsonlite', 'Matrix', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())"
+          RDscript${{ matrix.r_customization }} -e "install.packages(c('R6', 'data.table', 'jsonlite', 'knitr', 'Matrix', 'rmarkdown', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())"
           sh build-cran-package.sh --r-executable=RD${{ matrix.r_customization }}
           RD${{ matrix.r_customization }} CMD INSTALL lightgbm_*.tar.gz || exit -1
       - name: Run tests with sanitizers
@@ -225,7 +225,7 @@ jobs:
         shell: bash
         run: |
           export PATH=/opt/R-devel/bin/:${PATH}
-          Rscript -e "install.packages(c('R6', 'data.table', 'jsonlite', 'Matrix', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())"
+          Rscript -e "install.packages(c('R6', 'data.table', 'jsonlite', 'knitr', 'Matrix', 'rmarkdown', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())"
           sh build-cran-package.sh
           R CMD check --as-cran --run-donttest lightgbm_*.tar.gz || exit -1
           if grep -q -E "NOTE|WARNING|ERROR" lightgbm.Rcheck/00check.log; then

.github/workflows/r_valgrind.yml

@@ -7,7 +7,7 @@ on:
 jobs:
   test-r-valgrind:
     name: r-package (ubuntu-latest, R-devel, valgrind)
-    timeout-minutes: 180
+    timeout-minutes: 240
     runs-on: ubuntu-latest
     container: wch1/r-debug
     env:

.github/workflows/static_analysis.yml

@@ -57,7 +57,7 @@ jobs:
       - name: Install packages
         shell: bash
         run: |
-          Rscript -e "install.packages(c('R6', 'data.table', 'jsonlite', 'Matrix', 'roxygen2', 'testthat'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())"
+          Rscript -e "install.packages(c('R6', 'data.table', 'jsonlite', 'Matrix', 'roxygen2', 'testthat', 'knitr', 'rmarkdown'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())"
           sh build-cran-package.sh || exit -1
           R CMD INSTALL --with-keep.source lightgbm_*.tar.gz || exit -1
       - name: Test documentation

.vsts-ci.yml

@@ -19,7 +19,7 @@ resources:
     image: 'ubuntu:latest'
     options: "--name ci-container -v /usr/bin/docker:/tmp/docker:ro"
   - container: rbase
-    image: rocker/r-base
+    image: wch1/r-debug
 jobs:
 ###########################################
 - job: Linux
@@ -300,6 +300,7 @@ jobs:
   steps:
   - script: |
       LGB_VER=$(head -n 1 VERSION.txt | sed "s/rc/-/g")
+      Rscript -e "install.packages(c('R6', 'data.table', 'jsonlite', 'knitr', 'Matrix', 'rmarkdown', 'testthat'), dependencies = c('Depends', 'Imports', 'LinkingTo'), repos = 'https://cran.r-project.org', Ncpus = parallel::detectCores())" || exit -1
       sh build-cran-package.sh || exit -1
       mv lightgbm_${LGB_VER}.tar.gz $(Build.ArtifactStagingDirectory)/lightgbm-${LGB_VER}-r-cran.tar.gz
     displayName: 'Build CRAN R-package'

R-package/DESCRIPTION

@@ -45,8 +45,11 @@ URL: https://github.com/Microsoft/LightGBM
 BugReports: https://github.com/Microsoft/LightGBM/issues
 NeedsCompilation: yes
 Biarch: true
+VignetteBuilder: knitr
 Suggests:
+    knitr,
     processx,
+    rmarkdown,
     testthat
 Depends:
     R (>= 3.5),

R-package/README.md

@@ -152,6 +152,7 @@ Rscript build_r.R
 
 The `build_r.R` script builds the package in a temporary directory called `lightgbm_r`. It will destroy and recreate that directory each time you run the script. That script supports the following command-line options:
 
+- `--no-build-vignettes`: Skip building vignettes.
 - `--skip-install`: Build the package tarball, but do not install it.
 - `--use-gpu`: Build a GPU-enabled version of the library.
 - `--use-mingw`: Force the use of MinGW toolchain, regardless of R version.
@@ -266,6 +267,11 @@ sh build-cran-package.sh
 
 This will create a file `lightgbm_${VERSION}.tar.gz`, where `VERSION` is the version of `LightGBM`.
 
+That script supports the following command-line options:
+
+- `--no-build-vignettes`: Skip building vignettes.
+- `--r-executable=[path-to-executable]`: use an alternative build of R
+
 Also, CRAN package is generated with every commit to any repo's branch and can be found in "Artifacts" section of the associated Azure Pipelines run.
 
 ### Standard Installation from CRAN Package

R-package/pkgdown/_pkgdown.yml

@@ -41,6 +41,8 @@ navbar:
     href: ../
   - icon: fa-home fa-lg
     href: index.html
+  - text: Articles
+    href: articles/index.html
   - text: Reference
     href: reference/index.html
   right:

R-package/vignettes/basic_walkthrough.Rmd

@@ -0,0 +1,115 @@
+---
+title:
+  "Basic Walkthrough"
+description: >
+  This vignette describes how to train a LightGBM model for binary classification.
+output: rmarkdown::html_vignette
+vignette: >
+  %\VignetteIndexEntry{Basic Walkthrough}
+  %\VignetteEngine{knitr::rmarkdown}
+  %\VignetteEncoding{UTF-8}
+---
+
+```{r, include = FALSE}
+knitr::opts_chunk$set(
+  collapse = TRUE
+  , comment = "#>"
+  , warning = FALSE
+  , message = FALSE
+)
+```
+
+## Introduction
+
+Welcome to the world of [LightGBM](https://lightgbm.readthedocs.io/en/latest/), a highly efficient gradient boosting implementation (Ke et al. 2017).
+
+```{r setup}
+library(lightgbm)
+```
+
+This vignette will guide you through its basic usage. It will show how to build a simple binary classification model based on a subset of the `bank` dataset (Moro, Cortez, and Rita 2014). You will use the two input features "age" and "balance" to predict whether a client has subscribed a term deposit.
+
+## The dataset
+
+The dataset looks as follows.
+
+```{r}
+data(bank, package = "lightgbm")
+
+bank[1L:5L, c("y", "age", "balance")]
+
+# Distribution of the response
+table(bank$y)
+```
+
+## Training the model
+
+The R package of LightGBM offers two functions to train a model:
+
+- `lgb.train()`: This is the main training logic. It offers full flexibility but requires a `Dataset` object created by the `lgb.Dataset()` function.
+- `lightgbm()`: Simpler, but less flexible. Data can be passed without having to bother with `lgb.Dataset()`.
+
+### Using the `lightgbm()` function
+
+In a first step, you need to convert data to numeric. Afterwards, you are ready to fit the model by the `lightgbm()` function.
+
+```{r}
+# Numeric response and feature matrix
+y <- as.numeric(bank$y == "yes")
+X <- data.matrix(bank[, c("age", "balance")])
+
+# Train
+fit <- lightgbm(
+  data = X
+  , label = y
+  , num_leaves = 4L
+  , learning_rate = 1.0
+  , nrounds = 10L
+  , objective = "binary"
+  , verbose = -1L
+)
+
+# Result
+summary(predict(fit, X))
+```
+
+It seems to have worked! And the predictions are indeed probabilities between 0 and 1.
+
+### Using the `lgb.train()` function
+
+Alternatively, you can go for the more flexible interface `lgb.train()`. Here, as an additional step, you need to prepare `y` and `X` by the data API `lgb.Dataset()` of LightGBM. Parameters are passed to `lgb.train()` as a named list.
+
+```{r}
+# Data interface
+dtrain <- lgb.Dataset(X, label = y)
+
+# Parameters
+params <- list(
+  objective = "binary"
+  , num_leaves = 4L
+  , learning_rate = 1.0
+)
+
+# Train
+fit <- lgb.train(
+  params
+  , data = dtrain
+  , nrounds = 10L
+  , verbose = -1L
+)
+```
+
+Try it out! If stuck, visit LightGBM's [documentation](https://lightgbm.readthedocs.io/en/latest/R/index.html) for more details.
+
+```{r, echo = FALSE, results = "hide"}
+# Cleanup
+if (file.exists("lightgbm.model")) {
+  file.remove("lightgbm.model")
+}
+```
+
+## References
+
+Ke, Guolin, Qi Meng, Thomas Finley, Taifeng Wang, Wei Chen, Weidong Ma, Qiwei Ye, and Tie-Yan Liu. 2017. "LightGBM: A Highly Efficient Gradient Boosting Decision Tree." In Advances in Neural Information Processing Systems 30 (Nip 2017).
+
+Moro, Sérgio, Paulo Cortez, and Paulo Rita. 2014. "A Data-Driven Approach to Predict the Success of Bank Telemarketing." Decision Support Systems 62: 22–31.

build-cran-package.sh

@@ -11,23 +11,36 @@
 #                    non-standard builds of R, such as those provided in
 #                    https://github.com/wch/r-debug.
 #
+#     --no-build-vignettes Pass this flag to skip creating vignettes.
+#                          You might want to do this to avoid installing
+#                          vignette-only dependencies, or to avoid
+#                          portability issues.
+#
 # [usage]
 #
 #     # default usage
 #     sh build-cran-package.sh
 #
 #     # custom R build
 #     sh build-cran-package.sh --r-executable=RDvalgrind
+#
+#     # skip vignette building
+#     sh build-cran-package.sh --no-build-vignettes
 
 set -e
 
+# Default values of arguments
+BUILD_VIGNETTES=true
 LGB_R_EXECUTABLE=R
 
 while [ $# -gt 0 ]; do
   case "$1" in
     --r-executable=*)
       LGB_R_EXECUTABLE="${1#*=}"
       ;;
+    --no-build-vignettes=*)
+      BUILD_VIGNETTES=false
+      ;;
     *)
       echo "invalid argument '${1}'"
       exit -1
@@ -57,6 +70,10 @@ cp -R R-package/* "${TEMP_R_DIR}"
 cp -R include "${TEMP_R_DIR}/src/"
 cp -R src/* "${TEMP_R_DIR}/src/"
 
+if ${BUILD_VIGNETTES} ; then
+    cp docs/logo/LightGBM_logo_black_text.svg "${TEMP_R_DIR}/vignettes/"
+fi
+
 cp \
     external_libs/fast_double_parser/include/fast_double_parser.h \
     "${TEMP_R_DIR}/src/include/LightGBM"
@@ -169,8 +186,46 @@ cd "${TEMP_R_DIR}"
 
 cd "${ORIG_WD}"
 
-"${LGB_R_EXECUTABLE}" CMD build \
-    --keep-empty-dirs \
-    lightgbm_r
+if ${BUILD_VIGNETTES} ; then
+    "${LGB_R_EXECUTABLE}" CMD build \
+        --keep-empty-dirs \
+        lightgbm_r
+
+    echo "removing object files created by vignettes"
+    rm -rf ./_tmp
+    mkdir _tmp
+    TARBALL_NAME="lightgbm_${LGB_VERSION}.tar.gz"
+    mv "${TARBALL_NAME}" _tmp/
+
+    echo "untarring ${TARBALL_NAME}"
+    cd _tmp
+        tar -xvf "${TARBALL_NAME}" > /dev/null 2>&1
+        rm -rf "${TARBALL_NAME}"
+    cd ..
+    echo "done untarring ${TARBALL_NAME}"
+
+    echo "re-tarring ${TARBALL_NAME}"
+    tar \
+        -czv \
+        -C ./_tmp \
+        --exclude=*.a \
+        --exclude=*.dll \
+        --exclude=*.o \
+        --exclude=*.so \
+        --exclude=*.tar.gz \
+        --exclude=**/conftest.c \
+        --exclude=**/conftest.exe \
+        -f "${TARBALL_NAME}" \
+        lightgbm \
+    > /dev/null 2>&1
+    echo "Done creating ${TARBALL_NAME}"
+
+    rm -rf ./_tmp
+else
+    "${LGB_R_EXECUTABLE}" CMD build \
+        --keep-empty-dirs \
+        --no-build-vignettes \
+        lightgbm_r
+fi
 
 echo "Done building R package"