Rename data kwarg to observed in Model register_rv

Author: ricardoV94

pymc/model.py

@@ -1257,7 +1257,7 @@ def set_data(
         shared_object.set_value(values)
 
     def register_rv(
-        self, rv_var, name, data=None, total_size=None, dims=None, transform=UNSET, initval=None
+        self, rv_var, name, observed=None, total_size=None, dims=None, transform=UNSET, initval=None
     ):
         """Register an (un)observed random variable with the model.
 
@@ -1266,9 +1266,8 @@ def register_rv(
         rv_var: TensorVariable
         name: str
             Intended name for the model variable.
-        data: array_like (optional)
-            If data is provided, the variable is observed. If None,
-            the variable is unobserved.
+        observed: array_like (optional)
+            Data values for observed variables.
         total_size: scalar
             upscales logp of variable with ``coef = total_size/var.shape[0]``
         dims: tuple
@@ -1295,31 +1294,31 @@ def register_rv(
                 if dname not in self.dim_lengths:
                     self.add_coord(dname, values=None, length=rv_var.shape[d])
 
-        if data is None:
+        if observed is None:
             self.free_RVs.append(rv_var)
             self.create_value_var(rv_var, transform)
             self.add_random_variable(rv_var, dims)
             self.set_initval(rv_var, initval)
         else:
             if (
-                isinstance(data, Variable)
-                and not isinstance(data, (GenTensorVariable, Minibatch))
-                and data.owner is not None
+                isinstance(observed, Variable)
+                and not isinstance(observed, (GenTensorVariable, Minibatch))
+                and observed.owner is not None
                 # The only Aesara operation we allow on observed data is type casting
                 # Although we could allow for any graph that does not depend on other RVs
                 and not (
-                    isinstance(data.owner.op, Elemwise)
-                    and isinstance(data.owner.op.scalar_op, Cast)
+                    isinstance(observed.owner.op, Elemwise)
+                    and isinstance(observed.owner.op.scalar_op, Cast)
                 )
             ):
                 raise TypeError(
                     "Variables that depend on other nodes cannot be used for observed data."
-                    f"The data variable was: {data}"
+                    f"The data variable was: {observed}"
                 )
 
             # `rv_var` is potentially changed by `make_obs_var`,
             # for example into a new graph for imputation of missing data.
-            rv_var = self.make_obs_var(rv_var, data, dims, transform)
+            rv_var = self.make_obs_var(rv_var, observed, dims, transform)
 
         return rv_var
 

pymc/tests/test_model.py

@@ -1445,7 +1445,7 @@ def test_tag_future_warning_model():
         assert y is not x
         assert y.tag is not x.tag
         assert isinstance(y.tag, _FutureWarningValidatingScratchpad)
-        y = model.register_rv(y, name="y", data=5)
+        y = model.register_rv(y, name="y", observed=5)
         assert isinstance(y.tag, _FutureWarningValidatingScratchpad)
 
         # Test expected warnings

pymc/tests/test_model_graph.py

@@ -247,7 +247,7 @@ def model_non_random_variable_rvs():
 
         z_raw = pm.Normal.dist(y, shape=(5,))
         z = pm.math.clip(z_raw, -1, 1)
-        model.register_rv(z, name="z", data=[0] * 5)
+        model.register_rv(z, name="z", observed=[0] * 5)
 
     compute_graph = {
         "mu": set(),