Prior predictive guide (#1334)

* raw text guiding Claude

* init nb

* simple case init

* first iteration simple example

* add docs

* mmm init

* pre-commit

* allow prior_predictive

* mmm prior analysis

* mmm example code

* error

* intro mmm

* mmm text

* tv itercept priors

* prior intercept

* first final iteration

* rm file

* ab testing reference

* feedback part 1

* feedback p2

* stan link

* rm ugly cell output

docs/source/notebooks/index.md

@@ -46,4 +46,5 @@ customer_choice/mv_its_unsaturated
 
 general/model_configuration
 general/other_nuts_samplers
+general/prior_predictive
 :::

pymc_marketing/mmm/base.py

@@ -910,16 +910,24 @@ def plot_components_contributions(self, **plt_kwargs: Any) -> plt.Figure:
             )
         return fig
 
-    def compute_channel_contribution_original_scale(self) -> DataArray:
+    def compute_channel_contribution_original_scale(
+        self, prior: bool = False
+    ) -> DataArray:
         """Compute the channel contributions in the original scale of the target variable.
 
+        Parameters
+        ----------
+        prior : bool, optional
+            Whether to use the prior or posterior, by default False (posterior)
+
         Returns
         -------
         DataArray
 
         """
+        _data = self.prior if prior else self.fit_result
         channel_contribution = az.extract(
-            data=self.fit_result, var_names=["channel_contributions"], combined=False
+            data=_data, var_names=["channel_contributions"], combined=False
         )
 
         # sklearn preprocessers expect 2-D arrays of (obs, features)