Add _print_name to Truncated and CustomDists

pymc/distributions/distribution.py

@@ -663,6 +663,7 @@ def rv_op(
                 ndim_supp=ndim_supp,
                 ndims_params=ndims_params,
                 dtype=dtype,
+                _print_name=(class_name, f"\\operatorname{{{class_name}}}"),
                 # Specific to CustomDist
                 _random_fn=random,
             ),
@@ -802,6 +803,7 @@ def rv_op(
             # If logp is not provided, we try to infer it from the dist graph
             dict(
                 inline_logprob=logp is None,
+                _print_name=(class_name, f"\\operatorname{{{class_name}}}"),
             ),
         )
 

pymc/distributions/truncated.py

@@ -56,6 +56,10 @@ class TruncatedRV(SymbolicRandomVariable):
     def __init__(self, *args, base_rv_op: Op, max_n_steps: int, **kwargs):
         self.base_rv_op = base_rv_op
         self.max_n_steps = max_n_steps
+        self._print_name = (
+            f"Truncated{self.base_rv_op._print_name[0]}",
+            f"\\operatorname{{{self.base_rv_op._print_name[1]}}}",
+        )
         super().__init__(*args, **kwargs)
 
     def update(self, node: Node):

tests/test_printing.py

@@ -15,7 +15,7 @@
 
 from pytensor.tensor.random import normal
 
-from pymc import Bernoulli, Censored, HalfCauchy, Mixture, StudentT
+from pymc import Bernoulli, Censored, CustomDist, Gamma, HalfCauchy, Mixture, StudentT, Truncated
 from pymc.distributions import (
     Dirichlet,
     DirichletMultinomial,
@@ -285,3 +285,27 @@ def test_model_repr_variables_without_monkey_patched_repr():
 
     str_repr = model.str_repr()
     assert str_repr == "x ~ Normal(0, 1)"
+
+
+def test_truncated_repr():
+    with Model() as model:
+        x = Truncated("x", Gamma.dist(1, 1), lower=0, upper=20)
+
+    str_repr = model.str_repr(include_params=False)
+    assert str_repr == "x ~ TruncatedGamma"
+
+
+def test_custom_dist_repr():
+    with Model() as model:
+
+        def dist(mu, size):
+            return Normal.dist(mu, 1, size=size)
+
+        def random(rng, mu, size):
+            return rng.normal(mu, size=size)
+
+        x = CustomDist("x", 0, dist=dist, class_name="CustomDistNormal")
+        x = CustomDist("y", 0, random=random, class_name="CustomRandomNormal")
+
+    str_repr = model.str_repr(include_params=False)
+    assert str_repr == "\n".join(["x ~ CustomDistNormal", "y ~ CustomRandomNormal"])