Implement LATENT specification in cgpm_metamodel grammar, parser, and stub in _create_schema.\nProgress #435.

Author: F Saad

src/metamodels/cgpm_metamodel.py

@@ -648,6 +648,7 @@ def _create_schema(bdb, generator_id, schema_ast):
     # State.
     variables = []
     categoricals = {}
+    declared_latents = []
     cgpm_composition = []
     modelled = set()
     subsample = None
@@ -696,6 +697,18 @@ def _retrieve_stattype_dist_params(var):
             variables.append([var, stattype, dist, params])
             modelled.add(var)
 
+        elif isinstance(clause, cgpm_schema.parse.Latent):
+            # Reject if the latent variable has already been declared.
+            if any(l[0] == clause.name for l in declared_latents):
+                duplicate.add(clause.name)
+
+            # XXX FILL ME XXX
+
+            # Register the latent variable name and stattype into bayesdb.
+            # Do something related to the error checking data structures.
+
+            declared_latents.append((clause.name, clause.stattype))
+
         elif isinstance(clause, cgpm_schema.parse.Foreign):
             # Foreign model: some set of output variables is to be
             # modelled by foreign logic, possibly conditional on some
@@ -753,7 +766,7 @@ def _retrieve_stattype_dist_params(var):
             subsample = clause.n
 
         else:
-            assert False
+            raise BQLError(bdb, 'Unknown clause: %r' % (clause,))
 
     # Raise an exception if there were duplicates or unknown
     # variables.

src/metamodels/cgpm_schema/grammar.y

@@ -34,6 +34,7 @@ clause(basic)		::= var(var) dist(dist) param_opt(params).
 clause(foreign)		::= K_MODEL vars(outputs) given_opt(inputs)
 			K_USING foreign(name) param_opt(params).
 clause(subsamp)		::= K_SUBSAMPLE L_NUMBER(n).
+clause(latent)		::= K_LATENT var(var) stattype(st).
 
 dist(name)			::= L_NAME(dist).
 foreign(name)		::= L_NAME(foreign).
@@ -45,6 +46,8 @@ vars(one)			::= var(var).
 vars(many)			::= vars(vars) T_COMMA var(var).
 var(name)			::= L_NAME(var).
 
+stattype(s)			::= L_NAME(st).
+
 param_opt(none)		::= .
 param_opt(some)		::= T_LROUND params(ps) T_RROUND.
 params(one)			::= param(param).

src/metamodels/cgpm_schema/parse.py

@@ -31,6 +31,7 @@
 
 KEYWORDS = {
     'given': grammar.K_GIVEN,
+    'latent': grammar.K_LATENT,
     'model': grammar.K_MODEL,
     'subsample': grammar.K_SUBSAMPLE,
     'using': grammar.K_USING,
@@ -132,6 +133,8 @@ def p_clause_foreign(self, outputs, inputs, name, params):
         return Foreign(outputs, inputs, name, params)
     def p_clause_subsamp(self, n):
         return Subsample(n)
+    def p_clause_latent(self, var, st):
+        return Latent(var, st)
 
     def p_dist_name(self, dist):                return casefold(dist)
     def p_foreign_name(self, foreign):          return casefold(foreign)
@@ -143,6 +146,8 @@ def p_vars_one(self, var):                  return [var]
     def p_vars_many(self, vars, var):           vars.append(var); return vars
     def p_var_name(self, var):                  return var
 
+    def p_stattype_s(self, st):                   return st
+
     def p_param_opt_none(self):                 return []
     def p_param_opt_some(self, ps):             return ps
     def p_params_one(self, param):              return [param]
@@ -166,3 +171,8 @@ def p_param_nam(self, p, nam):              return (p, nam)
 Subsample = namedtuple('Subsample', [
     'n',
 ])
+
+Latent = namedtuple('Latent', [
+    'name',
+    'stattype',
+])

tests/foobar.py

@@ -152,6 +152,7 @@
         launch_mass NUMERICAL,
         perigee NUMERICAL,
         period NUMERICAL,
+
         LATENT kepler_noise NUMERICAL,
         LATENT kepler_cluster_id CATEGORICAL
         )
@@ -172,11 +173,18 @@
 bdb.execute('''
     CREATE GENERATOR g0 FOR satellites USING cgpm (
         apogee NORMAL,
+
+        LATENT kepler_cluster_id NUMERICAL,
+        LATENT kepler_noise NUMERICAL,
+
         MODEL kepler_cluster_id, kepler_noise, period GIVEN apogee, perigee
             USING venturescript (source = "{}"),
+
         MODEL perigee GIVEN apogee USING linreg,
+
         MODEL class_of_orbit GIVEN apogee, period, perigee
             USING forest (k = 4),
+
         SUBSAMPLE 100,
         )
     '''.format(kepler_source))