Validator: check datasetId in visualizations table exists in measurements table (#255)

Closes #229

Author: dweindl

petab/visualize/lint.py

@@ -127,6 +127,20 @@ def validate_visualization_df(problem: Problem) -> bool:
                     )
                     errors = True
 
+    if problem.measurement_df is not None:
+        referenced_datasets = set(filter(bool, vis_df[C.DATASET_ID].unique()))
+        if referenced_datasets:
+            existing_datasets = set(
+                filter(bool, problem.measurement_df[C.DATASET_ID].unique())
+            )
+            if not referenced_datasets.issubset(existing_datasets):
+                logger.error(
+                    f"Visualization table references {C.DATASET_ID}(s) "
+                    f"{referenced_datasets - existing_datasets}, but no such "
+                    "dataset(s) exist in the measurement table."
+                )
+                errors = True
+
     return errors
 
 

tests/test_visualization.py

@@ -178,11 +178,16 @@ def test_visualization_with_vis_and_sim(
     simulation_file_Isensee,
     close_fig,
 ):
-    validate_visualization_df(
-        petab.Problem(
-            condition_df=petab.get_condition_df(condition_file_Isensee),
-            visualization_df=petab.get_visualization_df(vis_spec_file_Isensee),
+    assert (
+        validate_visualization_df(
+            petab.Problem(
+                condition_df=petab.get_condition_df(condition_file_Isensee),
+                visualization_df=petab.get_visualization_df(
+                    vis_spec_file_Isensee
+                ),
+            )
         )
+        is False
     )
     plot_with_vis_spec(
         vis_spec_file_Isensee,
@@ -577,3 +582,21 @@ def test_validate(vis_file, request):
     assert False is validate_visualization_df(
         petab.Problem(visualization_df=petab.get_visualization_df(vis_file))
     )
+
+
+def test_validate_visualization_missing_dataset(
+    condition_file_Isensee,
+    data_file_Isensee,
+    vis_spec_file_Isensee,
+    simulation_file_Isensee,
+):
+    petab_problem = petab.Problem(
+        condition_df=petab.get_condition_df(condition_file_Isensee),
+        measurement_df=petab.get_measurement_df(data_file_Isensee),
+        visualization_df=petab.get_visualization_df(vis_spec_file_Isensee),
+    )
+
+    assert validate_visualization_df(petab_problem) is False
+
+    petab_problem.visualization_df.loc[0, petab.DATASET_ID] = "missing_dataset"
+    assert validate_visualization_df(petab_problem) is True