for dataset in args.input_datasets:
instances += utils.load_flattened_data(dataset)
role_constraints = utils.get_role_constraints(roles, instances)
formulas = get_formulas(args.formula_file)
domains = get_domains(roles, instances)
# generates the markov logic network
mln = MLN(logic="FirstOrderLogic", grammar="PRACGrammar")
for domain, values in domains.items(): # domains
for value in values:
if len(value) > 1:
const = ''.join(value)
elif len(value) > 0:
const = value[0]
else:
const = "None"
mln.update_domain({domain: [const]})
for role in roles.keys(): # predicates
mln.predicate(Predicate(
role, [role + "_d!"])) # hard-functional constraints only
for formula in formulas: # formulas
formula_str = "0.0 "
for idx in range(len(formula)):
role = formula[idx]
if idx + 1 < len(formula):
formula_str += role + "(+?" + role[:2] + ")" + " ^ "
else:
formula_str += role + "(+?" + role[:2] + ")"
mln << formula_str
mln.write()
mln.tofile(args.output_mln)
print("The initial MLN has been written to '" + args.output_mln + "'.")
# r_i2e[triple[0]] = i2e[triple[0]]
# r_e2i[i2e[triple[0]]] = triple[0]
# if triple[2] not in r_i2e:
# r_i2e[triple[2]] = i2e[triple[2]]
# r_e2i[i2e[triple[2]]] = triple[2]
# if triple[1] not in r_i2r:
# r_i2r[triple[1]] = i2r[triple[1]]
# r_r2i[i2r[triple[1]]] = triple[1]
# generates the markov logic network
mln = MLN(logic="FirstOrderLogic", grammar="PRACGrammar")
# declares the constants
for ent in e2i.keys():
ent_type = ent.split("-")[-1]
if ent_type == "r":
mln.update_domain({"rooms": [ent]})
elif ent_type == "l":
mln.update_domain({"locations": [ent]})
elif ent_type == "o":
mln.update_domain({"objects": [ent]})
elif ent_type == "a":
mln.update_domain({"actions": [ent]})
elif ent_type == "s":
mln.update_domain({"states": [ent]})
else:
print("Error: Unknown entity type for domains!")
exit()
# declares the predicates
rel2mln = {
"HasEffect": [["actions", "states"], ["IsAction(?x)", "IsState(?y)"]],
"InverseActionOf": [["actions", "actions"],
