ENH: Testing.

.github/workflows/test.yml

@@ -0,0 +1,72 @@
+name: R-CMD-check ANTsRNet
+
+on:
+  push:
+    branches: [ main, master ]
+  pull_request:
+    branches: [ main, master ]
+
+jobs:
+  test:
+    runs-on: ubuntu-22.04
+
+    env:
+      R_REMOTES_NO_ERRORS_FROM_WARNINGS: true
+      RETICULATE_PYTHON: /usr/bin/python3
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Setup R
+        uses: r-lib/actions/setup-r@v2
+
+      - name: Install system dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y libcurl4-openssl-dev libssl-dev libxml2-dev python3 python3-pip
+
+      - name: Cache R packages
+        uses: actions/cache@v3
+        with:
+          path: ~/.cache/R
+          key: ${{ runner.os }}-r-pkgs-${{ hashFiles('DESCRIPTION') }}
+          restore-keys: |
+            ${{ runner.os }}-r-pkgs-
+
+      - name: Install Python deps
+        run: |
+          python3 -m pip install --upgrade pip
+          python3 -m pip install tensorflow==2.12 numpy==1.23 h5py
+
+      - name: Install R dependencies with pinned reticulate
+        run: |
+          Rscript -e '
+            cat("✅ Installing R packages...\n");
+            install.packages(c("remotes", "testthat", "tensorflow", "reticulate"), repos = "https://cloud.r-project.org");
+            cat("✅ Pointing reticulate to system Python...\n");
+            reticulate::use_python(Sys.getenv("RETICULATE_PYTHON"), required = TRUE);
+            cat("✅ Installing TensorFlow via R package...\n");
+            tensorflow::install_tensorflow(version = "2.12");
+            '
+
+      - name: Configure GitHub PAT
+        run: |
+          echo "GITHUB_PAT=${{ secrets.GH_PAT }}" >> $GITHUB_ENV
+          echo "GITHUB_PAT=${{ secrets.GH_PAT }}" >> ~/.Renviron
+
+      - name: Install ANTsR
+        run: |
+          Rscript -e 'remotes::install_github("ANTsX/ANTsR", dependencies = TRUE)'
+
+      - name: Install ANTsRNet
+        run: |
+          Rscript -e 'remotes::install_local(".", dependencies = TRUE)'
+
+      - name: Run tests individually (non-fatal)
+        continue-on-error: true
+        run: |
+          for f in tests/testthat/test_*.R; do
+            echo "===== Running $f ====="
+            Rscript -e "testthat::test_file('$f', reporter = 'summary')" || echo "Test $f failed"
+          done

R/getPretrainedNetwork.R

@@ -86,6 +86,7 @@ getPretrainedNetwork <- function(
               "dktOuterWithSpatialPriors",
               "DesikanKillianyTourvilleOuter",
               "HarvardOxfordAtlasSubcortical",
+              "elBicho",
               "ex5_coronal_weights",
               "ex5_sagittal_weights",
               "allen_brain_mask_weights",

tests/testthat/test_brain.R

@@ -0,0 +1,65 @@
+# test_that("deepAtropos runs with default input", {
+# #   skip_if_not_installed("ANTsRNet")
+#   skip_if_not_installed("ANTsR")
+#   library(ANTsRNet); library(ANTsR)
+
+#   t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+#   seg <- deepAtropos(t1, verbose = FALSE)
+
+#   expect_type(seg, "list")
+#   expect_s4_class(seg$segmentationImage, "antsImage")
+# })
+
+# test_that("deepAtropos accepts list input with NULLs", {
+# #   skip_if_not_installed("ANTsRNet")
+#   skip_if_not_installed("ANTsR")
+#   library(ANTsRNet); library(ANTsR)
+
+#   t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+#   seg <- deepAtropos(list(t1, NULL, NULL), verbose = FALSE)
+
+#   expect_type(seg, "list")
+#   expect_s4_class(seg$segmentationImage, "antsImage")
+# })
+
+# test_that("DKT labeling versions 0 and 1 work", {
+# #   skip_if_not_installed("ANTsRNet")
+#   skip_if_not_installed("ANTsR")
+#   library(ANTsRNet); library(ANTsR)
+
+#   t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+#   dkt0 <- desikanKillianyTourvilleLabeling(t1, version = 0)
+#   dkt1 <- desikanKillianyTourvilleLabeling(t1, version = 1)
+
+#   expect_s4_class(dkt0, "antsImage")
+#   expect_s4_class(dkt1, "antsImage")
+# })
+
+# test_that("Harvard-Oxford Atlas labeling works", {
+# #   skip_if_not_installed("ANTsRNet")
+#   skip_if_not_installed("ANTsR")
+#   library(ANTsRNet); library(ANTsR)
+
+#   t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+#   hoa <- harvardOxfordAtlasLabeling(t1)
+# })
+
+test_that("deepFlash returns expected segmentation", {
+  skip_if_not_installed("ANTsRNet")
+  skip_if_not_installed("ANTsR")
+  library(ANTsRNet); library(ANTsR)
+
+  t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+  df <- deepFlash(t1, verbose = FALSE)
+  expect_type(df, "list")
+})
+
+test_that("claustrum segmentation runs", {
+  skip_if_not_installed("ANTsRNet")
+  skip_if_not_installed("ANTsR")
+  library(ANTsRNet); library(ANTsR)
+
+  t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+  seg <- claustrumSegmentation(t1)
+  expect_s4_class(seg, "antsImage")
+})

tests/testthat/test_brainExtraction.R

@@ -0,0 +1,26 @@
+test_that("brainExtraction runs correctly across modalities", {
+  skip_if_not_installed("ANTsRNet")
+  skip_if_not_installed("ANTsR")
+
+  library(ANTsR)
+  library(ANTsRNet)
+
+  # Download and read test image
+  t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+
+  # Define modalities to test
+  # modalities <- c("t1", "t1threetissue", "t1hemi", "t1lobes")
+  modalities <- c("t1")
+
+  for (mod in modalities) {
+    bext <- brainExtraction(t1, modality = mod, verbose = FALSE)
+
+    if (mod %in% c("t1")) {
+      expect_s4_class(bext, "antsImage")
+    } else {
+      expect_type(bext, "list")
+      expect_true("segmentationImage" %in% names(bext))
+      expect_s4_class(bext$segmentationImage, "antsImage")
+    }
+  }
+})

tests/testthat/test_generalApplications.R

@@ -0,0 +1,40 @@
+test_that("MRI super-resolution runs and returns antsImage", {
+  skip_if_not_installed("ANTsRNet")
+  skip_if_not_installed("ANTsR")
+
+  library(ANTsRNet)
+  library(ANTsR)
+
+  t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+  t1_lr <- resampleImage(t1, c(4, 4, 4), useVoxels = FALSE)
+  t1_sr <- mriSuperResolution(t1_lr, expansionFactor = c(1, 1, 2))
+  expect_s4_class(t1_sr, "antsImage")
+})
+
+test_that("T1w neural image QA returns numeric score", {
+  skip_if_not_installed("ANTsRNet")
+  skip_if_not_installed("ANTsR")
+
+  library(ANTsRNet)
+  library(ANTsR)
+
+  t1 <- antsImageRead(getANTsXNetData('mprage_hippmapp3r'))
+  qa_score <- tidNeuralImageAssessment(t1)
+  expect_type(qa_score, "list")
+})
+
+test_that("PSNR and SSIM return valid similarity values", {
+  skip_if_not_installed("ANTsRNet")
+  skip_if_not_installed("ANTsR")
+
+  library(ANTsRNet)
+  library(ANTsR)
+
+  r16 <- antsImageRead(getANTsRData("r16"))
+  r64 <- antsImageRead(getANTsRData("r64"))
+
+  psnr_val <- PSNR(r16, r64)
+  expect_equal(psnr_val, 10.37418, tolerance = 1e-3)
+  ssim_val <- SSIM(r16, r64)
+  expect_equal(ssim_val, 0.5654819, tolerance = 1e-3)
+})