Allow posterior prediction to extend state-centered temporal processes beyond the fitted time horizon.

AGENTS.md

@@ -0,0 +1,16 @@
+Functions, classes, and packages designs should be straightforward.
+The code needs to be clean and human-readable and human-maintainable.
+Factor repeated computation into general functions.
+Avoid overly general and extensible frameworks.
+Always ask if a design or code change requires backwards compatibility.
+In most cases, backwards compatibility is not a concern.
+
+coding style: always use MyPy type hints.
+all functions should have a docstring; public-facing API functions should be user-focussed.
+users and developers want to know what the function does, not design details.
+inline comments only for critical, non-obvious details.
+
+Design and implementation plans should be written in plain style.
+describe math in terms of python computation.
+
+I will commit and push edited files to GitHub myself - do not commit or push unless I tell you to.

pyrenew/latent/__init__.py

@@ -33,9 +33,6 @@
     AR1,
     DifferencedAR1,
     RandomWalk,
-    StateAR1,
-    StateDifferencedAR1,
-    StateRandomWalk,
     StepwiseTemporalProcess,
     TemporalProcess,
     WeeklyTemporalProcess,
@@ -69,9 +66,6 @@
     "AR1",
     "DifferencedAR1",
     "RandomWalk",
-    "StateAR1",
-    "StateDifferencedAR1",
-    "StateRandomWalk",
     "StepwiseTemporalProcess",
     "WeeklyTemporalProcess",
 ]