def test_distributions():
# test that the distributions come out right
# XXX: test more distributions
bandit = base.Bandit({
'loss':
hp_loguniform('lu', -2, 2) +
hp_qloguniform('qlu', np.log(1 + 0.01), np.log(20), 2) +
hp_quniform('qu', -4.999, 5, 1) + hp_uniform('u', 0, 10)
})
algo = base.Random(bandit)
trials = base.Trials()
exp = base.Experiment(trials, algo)
exp.catch_bandit_exceptions = False
N = 1000
exp.run(N)
assert len(trials) == N
idxs, vals = base.miscs_to_idxs_vals(trials.miscs)
print idxs.keys()
COUNTMAX = 130
COUNTMIN = 70
# -- loguniform
log_lu = np.log(vals['lu'])
assert len(log_lu) == N
assert -2 < np.min(log_lu)
assert np.max(log_lu) < 2
h = np.histogram(log_lu)[0]
print h
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
# -- quantized log uniform
qlu = vals['qlu']
assert np.all(np.fmod(qlu, 2) == 0)
assert np.min(qlu) == 2
assert np.max(qlu) == 20
bc_qlu = np.bincount(qlu)
assert bc_qlu[2] > bc_qlu[4] > bc_qlu[6] > bc_qlu[8]
# -- quantized uniform
qu = vals['qu']
assert np.min(qu) == -5
assert np.max(qu) == 5
assert np.all(np.fmod(qu, 1) == 0)
bc_qu = np.bincount(np.asarray(qu).astype('int') + 5)
assert np.all(40 < bc_qu), bc_qu # XXX: how to get the distribution flat
# with new rounding rule?
assert np.all(bc_qu < 125), bc_qu
assert np.all(bc_qu < COUNTMAX)
# -- uniform
u = vals['u']
assert np.min(u) > 0
assert np.max(u) < 10
h = np.histogram(u)[0]
print h
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
def test_distributions():
# test that the distributions come out right
# XXX: test more distributions
bandit = base.Bandit({
'loss': hp_loguniform('lu', -2, 2) +
hp_qloguniform('qlu', np.log(1 + 0.01), np.log(20), 2) +
hp_quniform('qu', -4.999, 5, 1) +
hp_uniform('u', 0, 10)})
algo = base.Random(bandit)
trials = base.Trials()
exp = base.Experiment(trials, algo)
exp.catch_bandit_exceptions = False
N = 1000
exp.run(N)
assert len(trials) == N
idxs, vals = base.miscs_to_idxs_vals(trials.miscs)
print idxs.keys()
COUNTMAX = 130
COUNTMIN = 70
# -- loguniform
log_lu = np.log(vals['lu'])
assert len(log_lu) == N
assert -2 < np.min(log_lu)
assert np.max(log_lu) < 2
h = np.histogram(log_lu)[0]
print h
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
# -- quantized log uniform
qlu = vals['qlu']
assert np.all(np.fmod(qlu, 2) == 0)
assert np.min(qlu) == 2
assert np.max(qlu) == 20
bc_qlu = np.bincount(qlu)
assert bc_qlu[2] > bc_qlu[4] > bc_qlu[6] > bc_qlu[8]
# -- quantized uniform
qu = vals['qu']
assert np.min(qu) == -5
assert np.max(qu) == 5
assert np.all(np.fmod(qu, 1) == 0)
bc_qu = np.bincount(np.asarray(qu).astype('int') + 5)
assert np.all(40 < bc_qu), bc_qu # XXX: how to get the distribution flat
# with new rounding rule?
assert np.all(bc_qu < 125), bc_qu
assert np.all(bc_qu < COUNTMAX)
# -- uniform
u = vals['u']
assert np.min(u) > 0
assert np.max(u) < 10
h = np.histogram(u)[0]
print h
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
def many_dists():
a = hp_choice('a', [0, 1, 2])
b = hp_randint('b', 10)
c = hp_uniform('c', 4, 7)
d = hp_loguniform('d', -2, 0)
e = hp_quniform('e', 0, 10, 3)
f = hp_qloguniform('f', 0, 3, 2)
g = hp_normal('g', 4, 7)
h = hp_lognormal('h', -2, 2)
i = hp_qnormal('i', 0, 10, 2)
j = hp_qlognormal('j', 0, 2, 1)
z = a + b + c + d + e + f + g + h + i + j
return {'loss': scope.float(scope.log(1e-12 + z**2))}
def many_dists():
a=hp_choice('a', [0, 1, 2])
b=hp_randint('b', 10)
c=hp_uniform('c', 4, 7)
d=hp_loguniform('d', -2, 0)
e=hp_quniform('e', 0, 10, 3)
f=hp_qloguniform('f', 0, 3, 2)
g=hp_normal('g', 4, 7)
h=hp_lognormal('h', -2, 2)
i=hp_qnormal('i', 0, 10, 2)
j=hp_qlognormal('j', 0, 2, 1)
z = a + b + c + d + e + f + g + h + i + j
return {'loss': scope.float(scope.log(1e-12 + z ** 2))}
def test_distributions():
# test that the distributions come out right
# XXX: test more distributions
space = {
'loss': (
hp_loguniform('lu', -2, 2) +
hp_qloguniform('qlu', np.log(1 + 0.01), np.log(20), 2) +
hp_quniform('qu', -4.999, 5, 1) +
hp_uniform('u', 0, 10)),
'status': 'ok'}
trials = base.Trials()
N = 1000
fmin(lambda x: x,
space=space,
algo=rand.suggest,
trials=trials,
max_evals=N,
rstate=np.random.RandomState(124),
catch_eval_exceptions=False)
assert len(trials) == N
idxs, vals = base.miscs_to_idxs_vals(trials.miscs)
print(list(idxs.keys()))
COUNTMAX = 130
COUNTMIN = 70
# -- loguniform
log_lu = np.log(vals['lu'])
assert len(log_lu) == N
assert -2 < np.min(log_lu)
assert np.max(log_lu) < 2
h = np.histogram(log_lu)[0]
print(h)
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
# -- quantized log uniform
qlu = vals['qlu']
assert np.all(np.fmod(qlu, 2) == 0)
assert np.min(qlu) == 2
assert np.max(qlu) == 20
bc_qlu = np.bincount(qlu)
assert bc_qlu[2] > bc_qlu[4] > bc_qlu[6] > bc_qlu[8]
# -- quantized uniform
qu = vals['qu']
assert np.min(qu) == -5
assert np.max(qu) == 5
assert np.all(np.fmod(qu, 1) == 0)
bc_qu = np.bincount(np.asarray(qu).astype('int') + 5)
assert np.all(40 < bc_qu), bc_qu # XXX: how to get the distribution flat
# with new rounding rule?
assert np.all(bc_qu < 125), bc_qu
assert np.all(bc_qu < COUNTMAX)
# -- uniform
u = vals['u']
assert np.min(u) > 0
assert np.max(u) < 10
h = np.histogram(u)[0]
print(h)
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
def test_vectorize_config0():
p0 = hp_uniform('p0', 0, 1)
p1 = hp_loguniform('p1', 2, 3)
p2 = hp_choice('p2', [-1, p0])
p3 = hp_choice('p3', [-2, p1])
p4 = 1
p5 = [3, 4, p0]
p6 = hp_choice('p6', [-3, p1])
d = locals()
d['p1'] = None # -- don't sample p1 all the time, only if p3 says so
config = as_apply(d)
N = as_apply('N:TBA')
expr = config
expr_idxs = scope.range(N)
vh = VectorizeHelper(expr, expr_idxs, build=True)
vconfig = vh.v_expr
full_output = as_apply([vconfig, vh.idxs_by_label(), vh.vals_by_label()])
if 1:
print('=' * 80)
print('VECTORIZED')
print(full_output)
print('\n' * 1)
fo2 = replace_repeat_stochastic(full_output)
if 0:
print('=' * 80)
print('VECTORIZED STOCHASTIC')
print(fo2)
print('\n' * 1)
new_vc = recursive_set_rng_kwarg(
fo2,
as_apply(np.random.RandomState(1))
)
if 0:
print('=' * 80)
print('VECTORIZED STOCHASTIC WITH RNGS')
print(new_vc)
Nval = 10
foo, idxs, vals = rec_eval(new_vc, memo={N: Nval})
print('foo[0]', foo[0])
print('foo[1]', foo[1])
assert len(foo) == Nval
if 0: # XXX refresh these values to lock down sampler
assert foo[0] == {
'p0': 0.39676747423066994,
'p1': None,
'p2': 0.39676747423066994,
'p3': 2.1281244479293568,
'p4': 1,
'p5': (3, 4, 0.39676747423066994) }
assert foo[1] != foo[2]
print(idxs)
print(vals['p3'])
print(vals['p6'])
print(idxs['p1'])
print(vals['p1'])
assert len(vals['p3']) == Nval
assert len(vals['p6']) == Nval
assert len(idxs['p1']) < Nval
p1d = dict(list(zip(idxs['p1'], vals['p1'])))
for ii, (p3v, p6v) in enumerate(zip(vals['p3'], vals['p6'])):
if p3v == p6v == 0:
assert ii not in idxs['p1']
if p3v:
assert foo[ii]['p3'] == p1d[ii]
if p6v:
print('p6', foo[ii]['p6'], p1d[ii])
assert foo[ii]['p6'] == p1d[ii]
def test_vectorize_config0():
p0 = hp_uniform('p0', 0, 1)
p1 = hp_loguniform('p1', 2, 3)
p2 = hp_choice('p2', [-1, p0])
p3 = hp_choice('p3', [-2, p1])
p4 = 1
p5 = [3, 4, p0]
p6 = hp_choice('p6', [-3, p1])
d = locals()
d['p1'] = None # -- don't sample p1 all the time, only if p3 says so
config = as_apply(d)
N = as_apply('N:TBA')
expr = config
expr_idxs = scope.range(N)
vh = VectorizeHelper(expr, expr_idxs, build=True)
vconfig = vh.v_expr
full_output = as_apply([vconfig, vh.idxs_by_label(), vh.vals_by_label()])
if 1:
print '=' * 80
print 'VECTORIZED'
print full_output
print '\n' * 1
fo2 = replace_repeat_stochastic(full_output)
if 0:
print '=' * 80
print 'VECTORIZED STOCHASTIC'
print fo2
print '\n' * 1
new_vc = recursive_set_rng_kwarg(fo2, as_apply(np.random.RandomState(1)))
if 0:
print '=' * 80
print 'VECTORIZED STOCHASTIC WITH RNGS'
print new_vc
Nval = 10
foo, idxs, vals = rec_eval(new_vc, memo={N: Nval})
print 'foo[0]', foo[0]
print 'foo[1]', foo[1]
assert len(foo) == Nval
if 0: # XXX refresh these values to lock down sampler
assert foo[0] == {
'p0': 0.39676747423066994,
'p1': None,
'p2': 0.39676747423066994,
'p3': 2.1281244479293568,
'p4': 1,
'p5': (3, 4, 0.39676747423066994)
}
assert foo[1] != foo[2]
print idxs
print vals['p3']
print vals['p6']
print idxs['p1']
print vals['p1']
assert len(vals['p3']) == Nval
assert len(vals['p6']) == Nval
assert len(idxs['p1']) < Nval
p1d = dict(zip(idxs['p1'], vals['p1']))
for ii, (p3v, p6v) in enumerate(zip(vals['p3'], vals['p6'])):
if p3v == p6v == 0:
assert ii not in idxs['p1']
if p3v:
assert foo[ii]['p3'] == p1d[ii]
if p6v:
print 'p6', foo[ii]['p6'], p1d[ii]
assert foo[ii]['p6'] == p1d[ii]
def test_distributions():
# test that the distributions come out right
# XXX: test more distributions
space = {
'loss': (hp_loguniform('lu', -2, 2) +
hp_qloguniform('qlu', np.log(1 + 0.01), np.log(20), 2) +
hp_quniform('qu', -4.999, 5, 1) + hp_uniform('u', 0, 10)),
'status':
'ok'
}
trials = base.Trials()
N = 1000
fmin(lambda x: x,
space=space,
algo=rand.suggest,
trials=trials,
max_evals=N,
rstate=np.random.RandomState(124),
catch_eval_exceptions=False)
assert len(trials) == N
idxs, vals = base.miscs_to_idxs_vals(trials.miscs)
print(list(idxs.keys()))
COUNTMAX = 130
COUNTMIN = 70
# -- loguniform
log_lu = np.log(vals['lu'])
assert len(log_lu) == N
assert -2 < np.min(log_lu)
assert np.max(log_lu) < 2
h = np.histogram(log_lu)[0]
print(h)
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
# -- quantized log uniform
qlu = vals['qlu']
assert np.all(np.fmod(qlu, 2) == 0)
assert np.min(qlu) == 2
assert np.max(qlu) == 20
bc_qlu = np.bincount(qlu)
assert bc_qlu[2] > bc_qlu[4] > bc_qlu[6] > bc_qlu[8]
# -- quantized uniform
qu = vals['qu']
assert np.min(qu) == -5
assert np.max(qu) == 5
assert np.all(np.fmod(qu, 1) == 0)
bc_qu = np.bincount(np.asarray(qu).astype('int') + 5)
assert np.all(40 < bc_qu), bc_qu # XXX: how to get the distribution flat
# with new rounding rule?
assert np.all(bc_qu < 125), bc_qu
assert np.all(bc_qu < COUNTMAX)
# -- uniform
u = vals['u']
assert np.min(u) > 0
assert np.max(u) < 10
h = np.histogram(u)[0]
print(h)
assert np.all(COUNTMIN < h)
assert np.all(h < COUNTMAX)
def logu_range(label, lower, upper):
"""Return a continuous replacement for one_of(.1, 1, 10)"""
return hp_loguniform(label, np.log(lower), np.log(upper))
def test_vectorize_config0():
p0 = hp_uniform("p0", 0, 1)
p1 = hp_loguniform("p1", 2, 3)
p2 = hp_choice("p2", [-1, p0])
p3 = hp_choice("p3", [-2, p1])
p4 = 1
p5 = [3, 4, p0]
p6 = hp_choice("p6", [-3, p1])
d = locals()
d["p1"] = None # -- don't sample p1 all the time, only if p3 says so
config = as_apply(d)
N = as_apply("N:TBA")
expr = config
expr_idxs = scope.range(N)
vh = VectorizeHelper(expr, expr_idxs, build=True)
vconfig = vh.v_expr
full_output = as_apply([vconfig, vh.idxs_by_label(), vh.vals_by_label()])
if 1:
print("=" * 80)
print("VECTORIZED")
print(full_output)
print("\n" * 1)
fo2 = replace_repeat_stochastic(full_output)
if 0:
print("=" * 80)
print("VECTORIZED STOCHASTIC")
print(fo2)
print("\n" * 1)
new_vc = recursive_set_rng_kwarg(fo2, as_apply(np.random.RandomState(1)))
if 0:
print("=" * 80)
print("VECTORIZED STOCHASTIC WITH RNGS")
print(new_vc)
Nval = 10
foo, idxs, vals = rec_eval(new_vc, memo={N: Nval})
print("foo[0]", foo[0])
print("foo[1]", foo[1])
assert len(foo) == Nval
if 0: # XXX refresh these values to lock down sampler
assert foo[0] == {
"p0": 0.39676747423066994,
"p1": None,
"p2": 0.39676747423066994,
"p3": 2.1281244479293568,
"p4": 1,
"p5": (3, 4, 0.39676747423066994),
}
assert (foo[1].keys() != foo[2].keys()) or (foo[1].values() !=
foo[2].values())
print(idxs)
print(vals["p3"])
print(vals["p6"])
print(idxs["p1"])
print(vals["p1"])
assert len(vals["p3"]) == Nval
assert len(vals["p6"]) == Nval
assert len(idxs["p1"]) < Nval
p1d = dict(list(zip(idxs["p1"], vals["p1"])))
for ii, (p3v, p6v) in enumerate(zip(vals["p3"], vals["p6"])):
if p3v == p6v == 0:
assert ii not in idxs["p1"]
if p3v:
assert foo[ii]["p3"] == p1d[ii]
if p6v:
print("p6", foo[ii]["p6"], p1d[ii])
assert foo[ii]["p6"] == p1d[ii]
