def test_args_chain_ids(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
jdata = os.path.join(datafiles_path, 'bernoulli.data.json')
args = SamplerArgs(model, chain_ids=[7, 9], data=jdata)
cmd = args.compose_command(0, 'output')
self.assertIn('bernoulli', cmd)
self.assertIn('bernoulli.data.json', cmd)
self.assertIn('id=7', cmd)
cmd = args.compose_command(1, 'output')
self.assertIn('id=9', cmd)
def test_args_many_1(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
jdata = os.path.join(datafiles_path, 'bernoulli.data.json')
jmetric = os.path.join(datafiles_path, 'bernoulli.metric.json')
output = os.path.join(TMPDIR, 'bernoulli.output')
args = SamplerArgs(model,
chain_ids=[1, 2],
seed=12345,
warmup_iters=100,
sampling_iters=100,
save_warmup=True,
thin=2,
metric=jmetric,
step_size=1.5,
data=jdata,
output_file=output,
max_treedepth=11,
adapt_delta=0.9)
cmd = args.compose_command(0, ''.join([output, '-1.csv']))
s1 = 'test/data/bernoulli id=1 random seed=12345 data file=test/data/bernoulli.data.json'
s2 = 'method=sample num_samples=100 num_warmup=100 save_warmup=1 thin=2'
s3 = 'algorithm=hmc engine=nuts max_depth=11 stepsize=1.5 metric=diag_e metric_file="test/data/bernoulli.metric.json" adapt delta=0.9'
self.assertIn(s1, cmd)
self.assertIn(s2, cmd)
self.assertIn(s3, cmd)
def test_args_step_size_2(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model, chain_ids=[1, 2], step_size=[1.31, 1.29])
cmd = args.compose_command(1, 'output')
self.assertIn('stepsize=1.29', cmd)
def test_args_adapt_delta_1(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model, chain_ids=[1, 2], adapt_delta=.93)
cmd = args.compose_command(0, 'output')
self.assertIn('adapt delta=0.93', cmd)
def test_args_save_warmup_2(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model, chain_ids=[1, 2], save_warmup=False)
cmd = args.compose_command(0, 'output')
self.assertNotIn('save_warmup', cmd)
def test_args_max_treedepth_good(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model, chain_ids=[1, 2], max_treedepth=15)
cmd = args.compose_command(0, 'output')
self.assertIn('max_depth=15', cmd)
def test_args_iters_2(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model, chain_ids=[1, 2], sampling_iters=123)
cmd = args.compose_command(0, 'output')
self.assertIn('num_samples=123', cmd)
def test_args_metric_4(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model, chain_ids=[1, 2], metric='dense_e')
cmd = args.compose_command(0, 'output')
s1 = 'metric=dense_e'
self.assertIn(s1, cmd)
def test_args_metric_file_1(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
jmetric = os.path.join(datafiles_path, 'bernoulli.metric.json')
args = SamplerArgs(model, chain_ids=[1, 2], metric=jmetric)
cmd = args.compose_command(0, 'output')
s1 = 'metric=diag_e metric_file="test/data/bernoulli.metric.json'
self.assertIn(s1, cmd)
def test_args_inits_3(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
jdata = os.path.join(datafiles_path, 'bernoulli.data.json')
args = SamplerArgs(model, chain_ids=[1, 2], data=jdata, inits=3.33)
cmd = args.compose_command(0, 'output')
s1 = 'data file=test/data/bernoulli.data.json init=3.33'
self.assertIn(s1, cmd)
def test_args_min(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
output = os.path.join(TMPDIR, 'bernoulli.output')
args = SamplerArgs(model, chain_ids=[1, 2], output_file=output)
args.validate()
cmd = args.compose_command(0, ''.join([output, '-1.csv']))
self.assertIn('id=1', cmd)
def test_args_warmup_schedule_1(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
args = SamplerArgs(model,
chain_ids=[1, 2],
warmup_iters=200,
warmup_schedule=(0.1, 0.8, 0.1))
cmd = args.compose_command(0, 'output')
s1 = 'algorithm=hmc adapt init_buffer=20 term_buffer=20'
def test_args_num_iters(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
output = os.path.join(TMPDIR, 'bernoulli.output')
args = SamplerArgs(model,
chain_ids=[1, 2],
output_file=output,
sampling_iters=3,
warmup_iters=7)
cmd = args.compose_command(0, ''.join([output, '-1.csv']))
self.assertIn('num_samples=3', cmd)
self.assertIn('num_warmup=7', cmd)
def test_args_good(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
rdata = os.path.join(datafiles_path, 'bernoulli.data.R')
output = os.path.join(TMPDIR, 'bernoulli.output')
args = SamplerArgs(model,
chain_ids=[1, 2],
seed=12345,
data=rdata,
output_file=output,
max_treedepth=15,
adapt_delta=0.99)
cmd = args.compose_command(0, ''.join([output, '-1.csv']))
self.assertIn('random seed=12345', cmd)
self.assertIn('data file=', cmd)
self.assertIn(
'algorithm=hmc engine=nuts max_depth=15 adapt delta=0.99', cmd)
def test_args_typical(self):
stan = os.path.join(datafiles_path, 'bernoulli.stan')
exe = os.path.join(datafiles_path, 'bernoulli')
model = Model(exe_file=exe, stan_file=stan)
jdata = os.path.join(datafiles_path, 'bernoulli.data.json')
output = os.path.join(TMPDIR, 'bernoulli.output')
args = SamplerArgs(model,
chain_ids=[1, 2],
seed=12345,
sampling_iters=100,
data=jdata,
output_file=output,
max_treedepth=11,
adapt_delta=0.9)
cmd = args.compose_command(0, ''.join([output, '-1.csv']))
self.assertIn('bernoulli', cmd)
self.assertIn('seed=12345', cmd)
self.assertIn('num_samples=100', cmd)
self.assertIn('bernoulli.data.json', cmd)
self.assertIn('algorithm=hmc engine=nuts max_depth=11 adapt delta=0.9',
cmd)
