if iter % 200 == 0 and iter != 0:
print rank, '---->', iter
fff()
Z, weighted_score, score, s, rec = probe(best_hypotheses, evaluation_data, pr_data, language.estimate_precision_and_recall)
to_file([[iter, Z, weighted_score, score, s, rec]], name)
if iter % options.STEPS == 0 and iter != 0:
size += 12
if staged: length += 4 * (size % 48 == 0)
for e in sampler.chains: e.data = get_data(n=size, max_length=length)
iter += 1
if __name__ == '__main__':
# ========================================================================================================
# Process command line arguments
# ========================================================================================================
(options, args) = parser.parse_args()
suffix = time.strftime('_' + options.NAME + '_%m%d_%H%M%S', time.localtime())
prefix = '../out/simulations/staged/'
# ========================================================================================================
# Running
# ========================================================================================================
language = AnBn()
show_info('running staged input case..')
probe_sampler(lambda: make_hypothesis('AnBn', N=options.N), language, options, prefix + 'staged_out_' + str(rank) + suffix, length=4)
show_info('running normal input case..')
probe_sampler(lambda: make_hypothesis('AnBn', N=options.N), language, options, prefix + 'normal_out_' + str(rank) + suffix)
# ========================================================================================================
# Process command line arguments
# ========================================================================================================
(options, args) = parser.parse_args()
suffix = time.strftime('_' + options.NAME + '_%m%d_%H%M%S',
time.localtime())
prefix = '../out/simulations/transfer/'
# ========================================================================================================
# Running
# ========================================================================================================
language = An()
show_info('running normal input case..')
sampler = probe_MHsampler(make_hypothesis('An', terminals=['b']),
language,
options,
prefix + 'without_prior_out_' + str(rank) +
suffix,
ret_sampler=True)
show_info('running with input using different letter case..')
CASE += 1
language = An(atom='b')
probe_MHsampler(make_hypothesis('An', terminals=['b']),
language,
options,
prefix + 'with_prior_out_' + str(rank) + suffix,
sampler=sampler)
In this case, we investigate the effect of different observed data distributions on training convergence.
"""
if __name__ == '__main__':
comm = MPI.COMM_WORLD
rank = comm.Get_rank()
# ========================================================================================================
# Process command line arguments
# ========================================================================================================
(options, args) = parser.parse_args()
suffix = time.strftime('_' + options.NAME + '_%m%d_%H%M%S', time.localtime())
prefix = '../out/simulations/skewed/'
# ========================================================================================================
# Running
# ========================================================================================================
language = AnBn()
show_info('running skewed input case..')
rec = probe_MHsampler(make_hypothesis('AnBn'), language, options, prefix + 'skewed_out_' + str(rank) + suffix)
show_info('running normal input case..')
CASE += 1
cnt = Counter()
num = 64.0 * 2 / options.FINITE
for i in xrange(1, options.FINITE/2+1):
cnt['a'*i+'b'*i] = num
rec1 = probe_MHsampler(make_hypothesis('AnBn'), language, options, prefix + 'normal_out' + str(rank) + suffix, data=[FunctionData(input=[], output=cnt)])
from LOTlib.Projects.FormalLanguageTheory.Model.Hypothesis import make_hypothesis
from LOTlib.Projects.FormalLanguageTheory.Language.An import An
import time
from mpi4py import MPI
register_primitive(flatten2str)
if __name__ == '__main__':
comm = MPI.COMM_WORLD
rank = comm.Get_rank()
# ========================================================================================================
# Process command line arguments
# ========================================================================================================
(options, args) = parser.parse_args()
suffix = time.strftime('_' + options.NAME + '_%m%d_%H%M%S', time.localtime())
prefix = '../out/simulations/transfer/'
# ========================================================================================================
# Running
# ========================================================================================================
language = An()
show_info('running normal input case..')
sampler = probe_MHsampler(make_hypothesis('An', terminals=['b']), language, options, prefix + 'without_prior_out_' + str(rank) + suffix, ret_sampler=True)
show_info('running with input using different letter case..')
CASE += 1
language = An(atom='b')
probe_MHsampler(make_hypothesis('An', terminals=['b']), language, options, prefix + 'with_prior_out_' + str(rank) + suffix, sampler=sampler)
# Process command line arguments
# ========================================================================================================
(options, args) = parser.parse_args()
suffix = time.strftime('_' + options.NAME + '_%m%d_%H%M%S',
time.localtime())
prefix = '../out/simulations/nonadjacent/'
# ========================================================================================================
# Running
# ========================================================================================================
# case 1
show_info('running predictable input case..')
language = LongDependency(C=['c'])
probe_MHsampler(
make_hypothesis('LongDependency', terminals=['c', 'd', 'e', 'f']),
language, options, prefix + 'c_long_' + str(rank) + suffix)
show_info('running predictable input case..')
CASE += 1
language = LongDependency(C=['c', 'd', 'e', 'f'])
probe_MHsampler(
make_hypothesis('LongDependency', terminals=['c', 'd', 'e', 'f']),
language, options, prefix + 'cdef_' + str(rank) + suffix)
# --------------------------------------------------------------------------------------------------------
# case 2
show_info('running predictable input case..')
options.FINITE = 4
CASE += 1
language = LongDependency(C=['c', 'd', 'e', 'f'])
# ========================================================================================================
# Process command line arguments
# ========================================================================================================
(options, args) = parser.parse_args()
suffix = time.strftime('_' + options.NAME + '_%m%d_%H%M%S', time.localtime())
prefix = '../out/simulations/nonadjacent/'
# ========================================================================================================
# Running
# ========================================================================================================
# case 1
show_info('running predictable input case..')
language = LongDependency(C=['c'])
probe_MHsampler(make_hypothesis('LongDependency', terminals=['c', 'd', 'e', 'f']), language, options, prefix + 'c_long_' + str(rank) + suffix)
show_info('running predictable input case..')
CASE += 1
language = LongDependency(C=['c', 'd', 'e', 'f'])
probe_MHsampler(make_hypothesis('LongDependency', terminals=['c', 'd', 'e', 'f']), language, options, prefix + 'cdef_' + str(rank) + suffix)
# --------------------------------------------------------------------------------------------------------
# case 2
show_info('running predictable input case..')
options.FINITE = 4
CASE += 1
language = LongDependency(C=['c', 'd', 'e', 'f'])
probe_MHsampler(make_hypothesis('LongDependency', terminals=['c', 'd', 'e', 'f']), language, options, prefix + 'c_short_' + str(rank) + suffix)