def makeObserves(ripl, num_points):
data = [
v.list([v.real(-2.4121470179012272),
v.real(-4.0898807104175177)]),
v.list([v.real(-3.5742473969392483),
v.real(16.047761595206289)]),
v.list([v.num(-0.44139814963230056),
v.num(-4.366209321011203)]),
v.list([v.num(-0.23258164982263235),
v.num(-5.687017265835844)]),
v.list([v.num(0.035565573263568961),
v.num(-3.337136380872824)]),
v.list([v.num(-4.3443008510924654),
v.num(17.827515632039944)]),
v.list([v.num(-0.19026889153157978),
v.num(-3.6065264508558972)]),
v.list([v.num(-4.8535236723040498),
v.num(17.098870618931819)]),
v.list([v.num(-1.1235421108040797),
v.num(-5.261098458581845)]),
v.list([v.num(1.1794444234381136),
v.num(-3.630199721409284)])
]
for i in range(num_points):
ripl.observe('(get_datapoint %d)' % i, data[i])
def testMoreElaborate(seed):
# Confirm that HMC still works in the presence of brush. Do not,
# however, mess with the possibility that the principal nodes that
# HMC operates over may themselves be in the brush.
ripl = get_ripl(seed=seed)
ripl.assume("x", "(tag (quote param) 0 (uniform_continuous -10 10))")
ripl.assume("y",
"(tag (quote param) 1 (uniform_continuous -10 10))",
label="pid")
ripl.assume("xout", """
(if (< x 0)
(normal x 1)
(normal x 2))""")
ripl.assume(
"out",
"(multivariate_normal (array xout y) (matrix (list (list 1 0.5) (list 0.5 1))))"
)
# Note: Can't observe coordinatewise because observe is not flexible
# enough. For this to work we would need observations of splits.
# ripl.observe("(lookup out 0)", 0)
# ripl.observe("(lookup out 1)", 0)
# Can't observe through the ripl literally because the string
# substitution (!) is not flexible enough.
# ripl.observe("out", [0, 0])
ripl.observe("out", val.list([val.real(0), val.real(0)]))
preds_mh = collectSamples(ripl,
"pid",
infer="(resimulation_mh default one 10)")
ripl.sivm.core_sivm.engine.reinit_inference_problem()
preds_hmc = collectSamples(ripl, "pid", infer="(hmc 'param all 0.1 20 10)")
return reportSameContinuous(preds_mh, preds_hmc)
def testMoveMatrix():
ripl = get_ripl()
ripl.assume("mu", "(array 0 0)")
ripl.assume("scale", "(matrix (list (list 2 1) (list 1 2)))")
ripl.assume("sigma", "(wishart scale 4)", label="pid")
ripl.assume("out", "(multivariate_normal mu sigma)")
ripl.observe("out", val.list([val.real(1), val.real(1)]))
preds_mh = collectSamples(ripl,
"pid",
infer="(resimulation_mh default one 30)")
ripl.sivm.core_sivm.engine.reinit_inference_problem()
preds_hmc = collectSamples(ripl, "pid", infer="(hmc default all 0.01 5 5)")
def test_observe(self):
inst = {
'instruction':'observe',
'expression': ['normal',v.number(1),v.number(2)],
'value': v.real(3)
}
o = self.sivm.execute_instruction(inst)
self.assertIsInstance(o['directive_id'],(int,float))
def test_force(self):
inst = {
'instruction': 'force',
'expression': ['normal', v.number(1),
v.number(2)],
'value': v.real(3)
}
self.sivm.execute_instruction(inst)
o2 = self.sivm.list_directives()
self.assertEquals(o2, [])
def test_observe_fail(self):
raise SkipTest("Engine should report a polite exception on constraint of a deterministic choice. Issue: https://app.asana.com/0/9277419963067/9940667562268")
inst = {
'instruction':'observe',
'expression': ['add',v.number(1),v.number(2)],
'value': v.real(4)
}
try:
self.sivm.execute_instruction(inst)
except VentureException as e:
self.assertEquals(e.exception, 'invalid_constraint')
def test_validate_value_1(self):
val = v.real(1)
self.assertEqual(utils.validate_value(val), val)
def pablo(nparticles):
# Example program from Github issue #571.
r = get_ripl(seed=1)
r.set_mode("venture_script")
r.execute_program("""
// Collapsed Dirichlet Process Mixture model using
// a Chinese Restaurant Process
// CRP and cluster
assume crp = make_crp(10);
assume get_cluster = mem(proc(id){
tag(quote(clustering), id, crp())
});
// Components
assume cmvn = mem(proc(cluster){
make_niw_normal(array(1, 1),1,2 + 1,matrix(array(array(1,0),array(0,1))))
});
// Sample data
assume get_datapoint = mem(proc(id){
cmvn(get_cluster(id))()
});
""")
r.observe("get_datapoint(1)",
vd.list([vd.real(12.4299342152),
vd.real(11.0243723622)]))
r.observe("get_datapoint(2)",
vd.list([vd.real(0.0876588636563),
vd.real(0.46296506961)]))
r.observe("get_datapoint(3)",
vd.list([vd.real(-7.99964564506),
vd.real(-8.98731353208)]))
r.observe("get_datapoint(4)",
vd.list([vd.real(8.23541122165),
vd.real(7.79102768844)]))
r.observe("get_datapoint(5)",
vd.list([vd.real(0.974810779937),
vd.real(1.38726814618)]))
r.observe("get_datapoint(6)",
vd.list([vd.real(9.01950900535),
vd.real(9.30739440906)]))
r.observe("get_datapoint(7)",
vd.list([vd.real(-10.5666520344),
vd.real(-9.79815099761)]))
r.observe("get_datapoint(8)",
vd.list([vd.real(-0.298512623055),
vd.real(-0.893928070534)]))
r.observe("get_datapoint(9)",
vd.list([vd.real(12.0800451026),
vd.real(7.67721242372)]))
r.observe("get_datapoint(10)",
vd.list([vd.real(8.77103714879),
vd.real(9.60268348563)]))
r.infer("resample(%d)" % (nparticles, ))
return r
