fix example

examples/howto_other_frameworks.md

@@ -52,15 +52,15 @@ print(sampling_fn())
 We do not care about the posterior distribution of the indicator variable `I_rv` so we marginalize it out, and subsequently build the logdensity's graph:
 
 ```{code-cell} python
-from aeppl import joint_logdensity
+from aeppl import joint_logprob
 
 y_vv = Y_rv.clone()
 i_vv = I_rv.clone()
 
 logdensity = []
 for i in range(4):
     i_vv = at.as_tensor(i, dtype="int64")
-    component_logdensity, _ = joint_logdensity(realized={Y_rv: y_vv, I_rv: i_vv})
+    component_logdensity, _ = joint_logprob(realized={Y_rv: y_vv, I_rv: i_vv})
     logdensity.append(component_logdensity)
 logdensity = at.stack(logdensity, axis=0)