Removed walrus operator, reverted bump to py3.8

CHANGELOG.md

@@ -1,7 +1,4 @@
-## [UNRELEASED] neptune-sklearn 2.2.0
-
-### Changes
-- Bumped minimum supported Python version to 3.8 ([#21](https://github.com/neptune-ai/neptune-sklearn/pull/21))
+## [UNRELEASED] neptune-sklearn 2.1.1
 
 ### Fixes
 - `create_*_summary()` now does not throw a `NeptuneUnsupportedType` error if expected metadata is not found ([#21](https://github.com/neptune-ai/neptune-sklearn/pull/21))

pyproject.toml

@@ -9,7 +9,7 @@ style = "semver"
 pattern = "default-unprefixed"
 
 [tool.poetry.dependencies]
-python = "^3.8"
+python = "^3.7"
 
 # Python lack of functionalities from future versions
 importlib-metadata = { version = "*", python = "<3.8" }

src/neptune_sklearn/impl/__init__.py

@@ -110,8 +110,8 @@ def create_regressor_summary(regressor, X_train, X_test, y_train, y_test, nrows=
         log_charts (`bool`, optional): Whether to calculate and log chart visualizations.
             Note: Calculating visualizations is potentially expensive depending on input data and regressor,
             and may take some time to finish. This is equivalent to calling the following functions from
-            this module: `create_learning_curve_chart()`, `create_feature_importance_chart()`, `create_residuals_chart()`,
-            `create_prediction_error_chart()`, and `create_cooks_distance_chart()`.
+            this module: `create_learning_curve_chart()`, `create_feature_importance_chart()`,
+            `create_residuals_chart()`, `create_prediction_error_chart()`, and `create_cooks_distance_chart()`.
 
     Returns:
         `dict` with all summary items.
@@ -147,15 +147,21 @@ def create_regressor_summary(regressor, X_train, X_test, y_train, y_test, nrows=
     reg_summary["integration/about/neptune-sklearn"] = __version__
 
     if log_charts:
-        if learning_curve := create_learning_curve_chart(regressor, X_train, y_train):
+        learning_curve = create_learning_curve_chart(regressor, X_train, y_train)
+        feature_importance = create_feature_importance_chart(regressor, X_train, y_train)
+        residuals = create_residuals_chart(regressor, X_train, X_test, y_train, y_test)
+        prediction_error = create_prediction_error_chart(regressor, X_train, X_test, y_train, y_test)
+        cooks_distance = create_cooks_distance_chart(regressor, X_train, y_train)
+
+        if learning_curve:
             reg_summary["diagnostics_charts/learning_curve"] = learning_curve
-        if feature_importance := create_feature_importance_chart(regressor, X_train, y_train):
+        if feature_importance:
             reg_summary["diagnostics_charts/feature_importance"] = feature_importance
-        if residuals := create_residuals_chart(regressor, X_train, X_test, y_train, y_test):
+        if residuals:
             reg_summary["diagnostics_charts/residuals"] = residuals
-        if prediction_error := create_prediction_error_chart(regressor, X_train, X_test, y_train, y_test):
+        if prediction_error:
             reg_summary["diagnostics_charts/prediction_error"] = prediction_error
-        if cooks_distance := create_cooks_distance_chart(regressor, X_train, y_train):
+        if cooks_distance:
             reg_summary["diagnostics_charts/cooks_distance"] = cooks_distance
 
     return reg_summary
@@ -223,15 +229,21 @@ def create_classifier_summary(classifier, X_train, X_test, y_train, y_test, nrow
     cls_summary["integration/about/neptune-sklearn"] = __version__
 
     if log_charts:
-        if classification_report := create_classification_report_chart(classifier, X_train, X_test, y_train, y_test):
+        classification_report = create_classification_report_chart(classifier, X_train, X_test, y_train, y_test)
+        confusion_matrix = create_confusion_matrix_chart(classifier, X_train, X_test, y_train, y_test)
+        ROC_AUC = create_roc_auc_chart(classifier, X_train, X_test, y_train, y_test)
+        precision_recall = create_precision_recall_chart(classifier, X_test, y_test)
+        class_prediction_error = create_class_prediction_error_chart(classifier, X_train, X_test, y_train, y_test)
+
+        if classification_report:
             cls_summary["diagnostics_charts/classification_report"] = classification_report
-        if confusion_matrix := create_confusion_matrix_chart(classifier, X_train, X_test, y_train, y_test):
+        if confusion_matrix:
             cls_summary["diagnostics_charts/confusion_matrix"] = confusion_matrix
-        if ROC_AUC := create_roc_auc_chart(classifier, X_train, X_test, y_train, y_test):
+        if ROC_AUC:
             cls_summary["diagnostics_charts/ROC_AUC"] = ROC_AUC
-        if precision_recall := create_precision_recall_chart(classifier, X_test, y_test):
+        if precision_recall:
             cls_summary["diagnostics_charts/precision_recall"] = precision_recall
-        if class_prediction_error := create_class_prediction_error_chart(classifier, X_train, X_test, y_train, y_test):
+        if class_prediction_error:
             cls_summary["diagnostics_charts/class_prediction_error"] = class_prediction_error
 
     return cls_summary
@@ -280,9 +292,12 @@ def create_kmeans_summary(model, X, nrows=1000, **kwargs):
     kmeans_summary["cluster_labels"] = get_cluster_labels(model, X, nrows=nrows, **kwargs)
     kmeans_summary["integration/about/neptune-sklearn"] = __version__
 
-    if kelbow := create_kelbow_chart(model, X, **kwargs):
+    kelbow = create_kelbow_chart(model, X, **kwargs)
+    silhouette = create_silhouette_chart(model, X, **kwargs)
+
+    if kelbow:
         kmeans_summary["diagnostics_charts/kelbow"] = kelbow
-    if silhouette := create_silhouette_chart(model, X, **kwargs):
+    if silhouette:
         kmeans_summary["diagnostics_charts/silhouette"] = silhouette
 
     return kmeans_summary