Merge branch 'main' into shared-multivariate

Author: AlexAndorra

notebooks/Structural Timeseries Modeling.ipynb

@@ -214,7 +214,7 @@ def make_symbolic_graph(self) -> None:
         self.ssm["selection", :, :] = pt.as_tensor_variable(selection_matrix)
 
         init_state = self.make_and_register_variable(
-            f"{self.name}",
+            f"params_{self.name}",
             shape=(k_endog_effective, 2) if k_endog_effective > 1 else (self.k_states,),
         )
         self.ssm["initial_state", :] = init_state.ravel()
@@ -264,26 +264,26 @@ def populate_component_properties(self):
                 for name in base_names
             ]
 
-        self.param_names = [f"{self.name}"]
+        self.param_names = [f"params_{self.name}"]
 
         if k_endog_effective == 1:
-            self.param_dims = {self.name: (f"state_{self.name}",)}
+            self.param_dims = {f"params_{self.name}": (f"state_{self.name}",)}
             self.coords = {f"state_{self.name}": base_names}
             self.param_info = {
-                f"{self.name}": {
+                f"params_{self.name}": {
                     "shape": (2,),
                     "constraints": None,
                     "dims": (f"state_{self.name}",),
                 }
             }
         else:
-            self.param_dims = {self.name: (f"endog_{self.name}", f"state_{self.name}")}
+            self.param_dims = {f"params_{self.name}": (f"endog_{self.name}", f"state_{self.name}")}
             self.coords = {
                 f"state_{self.name}": [f"Cos_{self.name}", f"Sin_{self.name}"],
                 f"endog_{self.name}": self.observed_state_names,
             }
             self.param_info = {
-                f"{self.name}": {
+                f"params_{self.name}": {
                     "shape": (k_endog_effective, 2),
                     "constraints": None,
                     "dims": (f"endog_{self.name}", f"state_{self.name}"),
@@ -295,7 +295,7 @@ def populate_component_properties(self):
             self.param_info[f"length_{self.name}"] = {
                 "shape": () if k_endog_effective == 1 else (k_endog_effective,),
                 "constraints": "Positive, non-zero",
-                "dims": None if k_endog_effective == 1 else f"endog_{self.name}",
+                "dims": None if k_endog_effective == 1 else (f"endog_{self.name}",),
             }
 
         if self.dampen:

pymc_extras/statespace/models/structural/components/cycle.py

@@ -22,7 +22,7 @@ def test_cycle_component_deterministic(rng):
     cycle = st.CycleComponent(
         name="cycle", cycle_length=12, estimate_cycle_length=False, innovations=False
     )
-    params = {"cycle": np.array([1.0, 1.0], dtype=config.floatX)}
+    params = {"params_cycle": np.array([1.0, 1.0], dtype=config.floatX)}
     x, y = simulate_from_numpy_model(cycle, rng, params, steps=12 * 12)
 
     assert_pattern_repeats(y, 12, atol=ATOL, rtol=RTOL)
@@ -32,7 +32,10 @@ def test_cycle_component_with_dampening(rng):
     cycle = st.CycleComponent(
         name="cycle", cycle_length=12, estimate_cycle_length=False, innovations=False, dampen=True
     )
-    params = {"cycle": np.array([10.0, 10.0], dtype=config.floatX), "dampening_factor_cycle": 0.75}
+    params = {
+        "params_cycle": np.array([10.0, 10.0], dtype=config.floatX),
+        "dampening_factor_cycle": 0.75,
+    }
     x, y = simulate_from_numpy_model(cycle, rng, params, steps=100)
 
     # check that cycle dampens to zero over time
@@ -44,7 +47,7 @@ def test_cycle_component_with_innovations_and_cycle_length(rng):
         name="cycle", estimate_cycle_length=True, innovations=True, dampen=True
     )
     params = {
-        "cycle": np.array([1.0, 1.0], dtype=config.floatX),
+        "params_cycle": np.array([1.0, 1.0], dtype=config.floatX),
         "length_cycle": 12.0,
         "dampening_factor_cycle": 0.95,
         "sigma_cycle": 1.0,
@@ -64,7 +67,7 @@ def test_cycle_multivariate_deterministic(rng):
         innovations=False,
         observed_state_names=["data_1", "data_2", "data_3"],
     )
-    params = {"cycle": np.array([[1.0, 1.0], [2.0, 2.0], [3.0, 3.0]], dtype=config.floatX)}
+    params = {"params_cycle": np.array([[1.0, 1.0], [2.0, 2.0], [3.0, 3.0]], dtype=config.floatX)}
     x, y = simulate_from_numpy_model(cycle, rng, params, steps=12 * 12)
 
     # Check that each variable has a cyclical pattern with the expected period
@@ -139,7 +142,7 @@ def test_cycle_multivariate_with_dampening(rng):
         observed_state_names=["data_1", "data_2", "data_3"],
     )
     params = {
-        "cycle": np.array([[10.0, 10.0], [20.0, 20.0], [30.0, 30.0]], dtype=config.floatX),
+        "params_cycle": np.array([[10.0, 10.0], [20.0, 20.0], [30.0, 30.0]], dtype=config.floatX),
         "dampening_factor_cycle": 0.75,
     }
     x, y = simulate_from_numpy_model(cycle, rng, params, steps=100)
@@ -167,7 +170,7 @@ def test_cycle_multivariate_with_innovations_and_cycle_length(rng):
         observed_state_names=["data_1", "data_2", "data_3"],
     )
     params = {
-        "cycle": np.array([[1.0, 1.0], [2.0, 2.0], [3.0, 3.0]], dtype=config.floatX),
+        "params_cycle": np.array([[1.0, 1.0], [2.0, 2.0], [3.0, 3.0]], dtype=config.floatX),
         "length_cycle": 12.0,
         "dampening_factor_cycle": 0.95,
         "sigma_cycle": np.array([0.5, 1.0, 1.5]),  # different innov variances per var

tests/statespace/models/structural/components/test_cycle.py

@@ -370,8 +370,8 @@ def create_structural_model_and_equivalent_statsmodel(
         params["length_cycle"] = cycle_length
 
         init_cycle = rng.normal(size=(2,)).astype(floatX)
-        params["cycle"] = init_cycle
-        expected_param_dims["cycle"] += ("state_cycle",)
+        params["params_cycle"] = init_cycle
+        expected_param_dims["params_cycle"] += ("state_cycle",)
 
         state_names = ["Cos_cycle", "Sin_cycle"]
         expected_coords["state_cycle"] += state_names