Requested changes

Author: cetagostini

docs/source/notebooks/mmm/mmm_multidimensional_example.ipynb

@@ -5142,7 +5142,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "If you have a custom model, you can wrappe it into the model protocol, and use the optimizer after. If your model handle scales internally, you don't need to modify anything. Otherwise, for the plots, you may want to use `scale_factor=N`. E.g:\n",
+    "If you have a custom model, you can wrapped it into the model protocol, and use the optimizer after. If your model handle scales internally, you don't need to modify anything. Otherwise, for the plots, you may want to use `scale_factor=N`. E.g:\n",
     "\n",
     "```python\n",
     "optimizable_model.plot.budget_allocation(\n",

pymc_marketing/mmm/utils.py

@@ -224,11 +224,11 @@ def create_zero_dataset(
     inferred_freq = pd.infer_freq(date_series.unique())
     if inferred_freq is None:  # fall-back if inference fails
         warnings.warn(
-            f"Could not infer frequency from '{date_col}'. Using daily ('D').",
+            f"Could not infer frequency from '{date_col}'. Using weekly ('W').",
             UserWarning,
             stacklevel=2,
         )
-        inferred_freq = "D"
+        inferred_freq = "W"
 
     # ---- 2. Build the full Cartesian product of dates X dims -------------------
     new_dates = pd.date_range(