Merge pull request #38 from fractal-napari-plugins-collection/improve…

…_testing

[WIP] Improve testing

.github/workflows/test_and_deploy.yaml

@@ -55,7 +55,7 @@ jobs:
         run: |
           pip install .
           # stop the build if there are Python syntax errors, undefined symbols or failed code style checks
-          python -m pylint --fail-under=8 -f parseable src
+          python -m pylint --ignore-patterns=test_.*.py --fail-under=8 -f parseable src
 
       # this runs the platform-specific tests declared in tox.ini
       - name: Test with tox

.pre-commit-config.yaml

@@ -0,0 +1,25 @@
+# enable pre-commit.ci at https://pre-commit.ci/
+# it adds:
+# 1. auto fixing pull requests
+# 2. auto updating the pre-commit configuration
+ci:
+  autoupdate_schedule: monthly
+  autofix_commit_msg: "style(pre-commit.ci): auto fixes [...]"
+  autoupdate_commit_msg: "ci(pre-commit.ci): autoupdate"
+
+repos:
+  - repo: https://github.com/abravalheri/validate-pyproject
+    rev: v0.12.1
+    hooks:
+      - id: validate-pyproject
+
+  - repo: https://github.com/charliermarsh/ruff-pre-commit
+    rev: v0.0.257
+    hooks:
+      - id: ruff
+        args: [--fix]
+
+  - repo: https://github.com/psf/black
+    rev: 23.1.0
+    hooks:
+      - id: black

setup.cfg

@@ -56,3 +56,7 @@ napari-feature-classifier =
 [options.entry_points]
 napari.manifest =
     napari-feature-classifier = napari_feature_classifier:napari.yaml
+
+[coverage:run]
+omit =
+    src/napari_feature_classifier/dev_main.py

src/napari_feature_classifier/_tests/test_annotator_widgets.py

@@ -1,14 +1,17 @@
 """ Tests for annotator widget initialization"""
-import numpy as np
 import imageio
 from pathlib import Path
-from napari_feature_classifier.annotator_init_widget import InitializeLabelAnnotatorWidget
+from napari_feature_classifier.annotator_init_widget import (
+    InitializeLabelAnnotatorWidget,
+)
 from napari_feature_classifier.annotator_widget import (
     LabelAnnotator,
-    get_class_selection
+    get_class_selection,
 )
 
-lbl_img_np = imageio.v2.imread(Path("src/napari_feature_classifier/sample_data/test_labels.tif"))
+lbl_img_np = imageio.v2.imread(
+    Path("src/napari_feature_classifier/sample_data/test_labels.tif")
+)
 
 
 # make_napari_viewer is a pytest fixture that returns a napari viewer object
@@ -22,15 +25,21 @@ def test_annotator_widgets(make_napari_viewer):
     label_layer = viewer.add_labels(lbl_img_np)
 
     # Start init widget
-    _ = InitializeLabelAnnotatorWidget(viewer)
+    init_widget = InitializeLabelAnnotatorWidget(viewer)
+    init_widget.initialize_annotator()
 
     # Start the annotator widget
     annotator_widget = LabelAnnotator(viewer)
 
     # call our widget method
-    #my_widget._on_click()
+    # my_widget._on_click()
     annotator_widget._init_annotation(label_layer)
 
+    # TODO: Test toggle_label
+
+
+# TODO: Test saving annotations
+
 
 def test_custom_class_selection(make_napari_viewer):
     """
@@ -40,12 +49,10 @@ def test_custom_class_selection(make_napari_viewer):
     viewer = make_napari_viewer()
     viewer.add_labels(lbl_img_np)
 
-    class_names = ['Class Test', 'Class XYZ', '12345']
+    class_names = ["Class Test", "Class XYZ", "12345"]
     # Start the annotator widget with a list of named classes
-    LabelAnnotator(
-        viewer,
-        ClassSelection = get_class_selection(class_names = class_names)
-    )
+    LabelAnnotator(viewer, ClassSelection=get_class_selection(class_names=class_names))
+
 
 def test_numbered_class_selection(make_napari_viewer):
     """
@@ -55,7 +62,4 @@ def test_numbered_class_selection(make_napari_viewer):
     viewer = make_napari_viewer()
     viewer.add_labels(lbl_img_np)
     # Start the annotator widget with a given number of classes
-    LabelAnnotator(
-        viewer,
-        ClassSelection = get_class_selection(n_classes = 8)
-    )
+    LabelAnnotator(viewer, ClassSelection=get_class_selection(n_classes=8))

src/napari_feature_classifier/_tests/test_classifier_widget.py

@@ -14,8 +14,13 @@
     LoadClassifierContainer,
 )
 
-lbl_img_np = imageio.v2.imread(Path("src/napari_feature_classifier/sample_data/test_labels.tif"))
-clf_path = Path("src/napari_feature_classifier/sample_data/test_labels_classifier.clf")
+lbl_img_np = imageio.v2.imread(
+    Path("src/napari_feature_classifier/sample_data/test_labels.tif")
+)
+clf_path = Path(
+    "src/napari_feature_classifier/sample_data/" "test_labels_classifier.clf"
+)
+
 
 # make_napari_viewer is a pytest fixture that returns a napari viewer object
 # capsys is a pytest fixture that captures stdout and stderr output streams
@@ -32,14 +37,21 @@ def test_classifier_widgets_initialization_no_features_selected(make_napari_view
 
     classifier_widget.initialize_run_widget()
 
-    # TODO: Catch that it doesn't actually initialize, but instead shows a 
+    # TODO: Catch that it doesn't actually initialize, but instead shows a
     # message to the user that now features were selected
     assert classifier_widget._run_container is None
 
 
+def test_classifier_initializtion_without_label_image(make_napari_viewer):
+    viewer = make_napari_viewer()
+    classifier_widget = ClassifierWidget(viewer)
+    assert classifier_widget._init_container._last_selected_label_layer is None
+
 
 features = make_features(np.unique(lbl_img_np)[1:], roi_id="ROI1", n_features=6)
 features_no_roi_id = features.drop(columns=["roi_id"])
+
+
 # make_napari_viewer is a pytest fixture that returns a napari viewer object
 @pytest.mark.parametrize("features", [features, features_no_roi_id])
 def test_running_classification_through_widget(features, make_napari_viewer):
@@ -85,9 +97,137 @@ def test_running_classification_through_widget(features, make_napari_viewer):
     # Delete the classifier file (cleanup to avoid overwriting confirmation)
     os.remove("lbl_img_np_classifier.clf")
 
+
+# Test classifier results export
+def test_prediction_export(make_napari_viewer, capsys):
+    """
+    Tests if the main widget launches
+    """
+    # make viewer and add an image layer using our fixture
+    viewer = make_napari_viewer()
+    label_layer = viewer.add_labels(lbl_img_np)
+    label_layer.features = features
+
+    # Start init widget
+    classifier_widget = ClassifierWidget(viewer)
+
+    # Select relevant features
+    classifier_widget._init_container._feature_combobox.value = [
+        "feature_1",
+        "feature_2",
+        "feature_3",
+    ]
+
+    classifier_widget.initialize_run_widget()
+
+    # Check that the run widget is initialized
+    assert classifier_widget._run_container is not None
+
+    # Add some annotations manually
+    label_layer.features.loc[0, "annotations"] = 1.0
+    label_layer.features.loc[1, "annotations"] = 1.0
+    label_layer.features.loc[3, "annotations"] = 3.0
+
+    # Run the classifier
+    classifier_widget._run_container.run()
+
+    # Test result export
+    classifier_widget._run_container.export_results()
+    df = pd.read_csv(classifier_widget._run_container._export_destination.value)
+    assert df.shape == (16, 5)
+    assert df["prediction"].isna().sum() == 0
+
+    # Delete the classifier file (cleanup to avoid overwriting confirmation)
+    os.remove("lbl_img_np_classifier.clf")
+    os.remove("lbl_img_np_predictions.csv")
+
+    message = "INFO: Annotations were saved at lbl_img_np_predictions.csv"
+    assert message in capsys.readouterr().out
+
+
 # TODO: Add a test to check the overwrite confirmations working correctly
 # For classifier files, for annotations and for exported predictions
 
+
+features = make_features(np.unique(lbl_img_np)[1:], roi_id="ROI1", n_features=6)
+
+
+# make_napari_viewer is a pytest fixture that returns a napari viewer object
+def test_classifier_fails_running_without_annotation(make_napari_viewer, capsys):
+    """
+    Tests if the main widget launches
+    """
+    # make viewer and add an image layer using our fixture
+    viewer = make_napari_viewer()
+    label_layer = viewer.add_labels(lbl_img_np)
+    label_layer.features = features
+
+    # Start init widget
+    classifier_widget = ClassifierWidget(viewer)
+
+    # Select relevant features
+    classifier_widget._init_container._feature_combobox.value = [
+        "feature_1",
+        "feature_2",
+        "feature_3",
+    ]
+
+    classifier_widget.initialize_run_widget()
+
+    # Run the classifier
+    classifier_widget._run_container.run()
+
+    expected_message = (
+        "INFO: Training failed. A typical reason are not "
+        "having enough annotations. \nThe error message was: Found array "
+        "with 0 sample(s) (shape=(0, 3)) while a minimum of 1 is required "
+        "by RandomForestClassifier."
+    )
+    assert expected_message in capsys.readouterr().out
+
+
+def test_layer_selection_changes(make_napari_viewer):
+    """
+    Tests if the main widget launches
+    """
+    # make viewer and add an image layer using our fixture
+    viewer = make_napari_viewer()
+    label_layer1 = viewer.add_labels(lbl_img_np, name="test_labels")
+    label_layer2 = viewer.add_labels(lbl_img_np, name="test_labels_2")
+    label_layer2.features = features
+    label_layer1.features = features
+
+    # Start init widget
+    classifier_widget = ClassifierWidget(viewer)
+
+    # Select relevant features
+    classifier_widget._init_container._feature_combobox.value = [
+        "feature_1",
+        "feature_2",
+        "feature_3",
+    ]
+
+    classifier_widget.initialize_run_widget()
+
+    assert (
+        classifier_widget._run_container._last_selected_label_layer.name
+        == "test_labels_2"
+    )
+    assert (
+        str(classifier_widget._run_container._export_destination.value)
+        == "test_labels_2_predictions.csv"
+    )
+    viewer.layers.selection.active = label_layer1
+    assert (
+        classifier_widget._run_container._last_selected_label_layer.name
+        == "test_labels"
+    )
+    assert (
+        classifier_widget._run_container._export_destination.value.name
+        == "test_labels_predictions.csv"
+    )
+
+
 # make_napari_viewer is a pytest fixture that returns a napari viewer object
 # capsys is a pytest fixture that captures stdout and stderr output streams
 def test_load_classifier_widget(make_napari_viewer, capsys):
@@ -109,3 +249,12 @@ def test_load_classifier_widget(make_napari_viewer, capsys):
     assert loading_widget._run_container._prediction_layer.visible
     assert "prediction" in label_layer.features.columns
 
+
+def test_change_of_loader_filter(make_napari_viewer, capsys):
+    viewer = make_napari_viewer()
+    loading_widget = LoadClassifierContainer(viewer)
+    assert loading_widget._filter.value == "*.clf"
+    assert loading_widget._clf_destination.filter is None
+    loading_widget._filter.value = "*.pkl"
+    assert loading_widget._clf_destination.filter == "*.pkl"
+    assert loading_widget._filter.value == "*.pkl"

src/napari_feature_classifier/_tests/test_feature_loading.py

@@ -0,0 +1,35 @@
+""" Tests feature loading"""
+import pandas as pd
+
+import imageio
+from pathlib import Path
+from napari_feature_classifier.feature_loader_widget import (
+    load_features_factory,
+)
+
+lbl_img_np = imageio.v2.imread(
+    Path("src/napari_feature_classifier/sample_data/test_labels.tif")
+)
+csv_path = Path("src/napari_feature_classifier/sample_data/test_df.csv")
+
+
+def test_feature_loading_csv(make_napari_viewer, capsys):
+    """
+    Tests if the main widget launches
+    """
+    # make viewer and add an image layer using our fixture
+    viewer = make_napari_viewer()
+    labels_layer = viewer.add_labels(lbl_img_np)
+
+    # Start feature loading widget
+    loading_widget = load_features_factory()
+    loading_widget.layer.value = labels_layer
+    loading_widget.path.value = csv_path
+    features = pd.read_csv(csv_path)
+    loading_widget.__call__()
+    assert (labels_layer.features == features).all().all()
+
+    # Assert that this message is logged
+    expected = "INFO: Loaded features and attached them "
+    expected += 'to "lbl_img_np" layer\n'
+    assert expected in capsys.readouterr().out

src/napari_feature_classifier/classifier_widget.py

@@ -17,7 +17,6 @@
     PushButton,
     Select,
 )
-from qtpy.QtWidgets import QMessageBox  # pylint: disable=E0611
 
 from napari_feature_classifier.annotator_init_widget import LabelAnnotatorTextSelector
 from napari_feature_classifier.annotator_widget import (
@@ -117,7 +116,9 @@ def get_class_names(self):
     def get_feature_options(self, layer):
         """
         Get the feature options of the currently selected layer
-        Only works if a label layer is selected (we don't load features from other layers)
+
+        Only works if a label layer is selected (we don't load features from
+        other layers)
         """
         if isinstance(layer, napari.layers.Labels):
             return list(layer.features.columns)
@@ -265,7 +266,7 @@ def __init__(
         # Export options
         self._export_destination = FileEdit(
             label="Prediction Export Path",
-            value=f"{self._last_selected_label_layer}_prediction.csv",
+            value=f"{self._last_selected_label_layer}_predictions.csv",
             mode="w",
         )
         self._export_button = PushButton(text="Export Classifier Result")
@@ -552,7 +553,7 @@ class LoadClassifierContainer(Container):
         The button that launches the `ClassifierRunContainer`
     _filter: magicgui.widgets.RadioButtons
         The radio button widget that allows the user to select the file filter
-        to use for selecting the classifier file. See 
+        to use for selecting the classifier file. See
         https://github.com/fractal-napari-plugins-collection/napari-feature-classifier/issues/36
         for more details.
     _run_container: ClassifierRunContainer