Generalize recipe for regularizations for n dims (#69)

Generalize the recipe to build an l2 regularization to mesh dimensions
other than 3.

Author: santisoler

src/inversion_ideas/recipes.py

@@ -307,9 +307,14 @@ def create_tikhonov_regularization(
     -----
     TODO
     """
-    # TODO: raise errors:
-    #   if dims == 2 and alpha_z is passed
-    #   if dims == 1 and alpha_y or alpha_z are passed
+    ndims = mesh.dim
+    if ndims == 2 and alpha_z is not None:
+        msg = f"Cannot pass 'alpha_z' when mesh has {ndims} dimensions."
+        raise TypeError(msg)
+    if ndims == 1 and (alpha_y is not None or alpha_z is not None):
+        msg = "Cannot pass 'alpha_y' nor 'alpha_z' when mesh has 1 dimension."
+        raise TypeError(msg)
+
     smallness = Smallness(
         mesh,
         active_cells=active_cells,
@@ -326,16 +331,24 @@ def create_tikhonov_regularization(
     if reference_model_in_flatness:
         kwargs["reference_model"] = reference_model
 
-    flatness_x = Flatness(mesh, **kwargs, direction="x")
-    if alpha_x is not None:
-        flatness_x = alpha_x * flatness_x
-
-    flatness_y = Flatness(mesh, **kwargs, direction="y")
-    if alpha_y is not None:
-        flatness_y = alpha_y * flatness_y
-
-    flatness_z = Flatness(mesh, **kwargs, direction="z")
-    if alpha_z is not None:
-        flatness_z = alpha_z * flatness_z
-
-    return (smallness + flatness_x + flatness_y + flatness_z).flatten()
+    match ndims:
+        case 3:
+            directions = ("x", "y", "z")
+            alphas = (alpha_x, alpha_y, alpha_z)
+        case 2:
+            directions = ("x", "y")
+            alphas = (alpha_x, alpha_y)
+        case 1:
+            directions = ("x",)
+            alphas = (alpha_x,)
+        case _:
+            raise ValueError()
+
+    regularization = smallness
+    for direction, alpha in zip(directions, alphas, strict=True):
+        phi = Flatness(mesh, **kwargs, direction=direction)
+        if alpha is not None:
+            phi = alpha * phi
+        regularization = regularization + phi
+
+    return regularization.flatten()