diff --git a/changelog_entry.yaml b/changelog_entry.yaml
index e69de29b..a07555d0 100644
--- a/changelog_entry.yaml
+++ b/changelog_entry.yaml
@@ -0,0 +1,4 @@
+- bump: patch
+  changes:
+    fixed:
+    - Fixed a bug in Core monthlyisation.
diff --git a/policyengine_core/data/dataset.py b/policyengine_core/data/dataset.py
index 05a2cba0..3235c956 100644
--- a/policyengine_core/data/dataset.py
+++ b/policyengine_core/data/dataset.py
@@ -33,7 +33,7 @@ class Dataset:
 
     _table_cache: Dict[str, pd.DataFrame] = None
 
-    def __init__(self, regenerate: bool = False):
+    def __init__(self, require: bool = False):
         # Setup dataset
         if self.file_path is None:
             raise ValueError(
@@ -66,9 +66,7 @@ def __init__(self, regenerate: bool = False):
                 "You tried to instantiate a Dataset object, but no time_period has been provided."
             )
 
-        if regenerate:
-            self.generate()
-        elif not self.exists:
+        if not self.exists and require:
             if self.url is not None:
                 self.download()
             else:
diff --git a/policyengine_core/simulations/simulation.py b/policyengine_core/simulations/simulation.py
index 40ec8b30..d68ecd28 100644
--- a/policyengine_core/simulations/simulation.py
+++ b/policyengine_core/simulations/simulation.py
@@ -153,7 +153,7 @@ def __init__(
                     )
                 dataset = datasets_by_name[dataset]
             if isinstance(dataset, type):
-                self.dataset: Dataset = dataset()
+                self.dataset: Dataset = dataset(require=True)
             else:
                 self.dataset = dataset
             self.build_from_dataset()
@@ -227,6 +227,7 @@ def build_from_dataset(self) -> None:
 
         builder = SimulationBuilder()
         builder.populations = self.populations
+
         try:
             data = self.dataset.load()
         except FileNotFoundError as e:
@@ -576,6 +577,7 @@ def _calculate(
                     start_instants = [
                         str(known_period.start)
                         for known_period in known_periods
+                        if known_period.unit == variable.definition_period
                     ]
                     latest_known_period = known_periods[
                         np.argmax(start_instants)