mrp(11,15) :- mrp(23).
mrp(X,Y) :- mrp(Y,X).
:- X=0..15, #count{Y: mrp(X,Y)} = 1.
:- X=0..15, #count{Y: mrp(X,Y)} >= 3.
reachable(X, Y) :- mrp(X, Y).
reachable(X, Y) :- reachable(X, Z), mrp(Z, Y).
:- mrp(X, _), mrp(Y, _), not reachable(X, Y).
'''
########
# Define nnMapping and initialze NeurASP object
########
m = FC(40, *[50, 50, 50, 50, 50], 24)
nnMapping = {'mrp': m}
NeurASPobj = NeurASP(dprogram+aspProgram, nnMapping, optimizers=None)
########
# Load pretrained model
########
saveModelPath = 'data/model.pt'
m.load_state_dict(torch.load(saveModelPath, map_location='cpu'))
########
# Start testing
########
NeurASPobj.testConstraint(dataList=dataListTest, obsList=obsListTest, mvppList=[aspProgram])
sp(9,13) :- nn_edge(g, 17, t).
sp(2,6) :- nn_edge(g, 18, t).
sp(6,10) :- nn_edge(g, 19, t).
sp(10,14) :- nn_edge(g, 20, t).
sp(3,7) :- nn_edge(g, 21, t).
sp(7,11) :- nn_edge(g, 22, t).
sp(11,15) :- nn_edge(g, 23, t).
sp(X,Y) :- sp(Y,X).
mistake :- X=0..15, #count{Y: sp(X,Y)} = 1.
mistake :- X=0..15, #count{Y: sp(X,Y)} >= 3.
reachable(X, Y) :- sp(X, Y).
reachable(X, Y) :- reachable(X, Z), sp(Z, Y).
mistake :- sp(X, _), sp(Y, _), not reachable(X, Y).
'''
m = FC(40, *[50, 50, 50, 50, 50], 24)
nnMapping = {'m': m}
optimizer = {'m':torch.optim.Adam(m.parameters(), lr=0.001)}
dlpmlnObj = DeepLPMLN(dprogram, nnMapping, optimizer)
# process the data
dataset = GridProbData("data/data.txt")
# print(dataset.valid_data.shape)
# print(dataset.train_labels[0])
# sys.exit()