def hyper_param(self):
data = get_data_from_s3(self._bucket, "miso/hyper/{0}_lrMU2".format(self.method_name))
return data['hyperparam']
### copy back from /backup the files where IS 1 is centered
directory_backup = '/miso/data/backup'
new_func_name = 'lrMU2'
#
# num_pts = 1000
# for index_IS in func.getList_IS_to_query():
# key = directory_backup+'/hyper_{1}_IS_{0}_{2}_points'.format(index_IS, func_name, num_pts)
# data = get_data_from_s3(bucket, key)
#
# # write to lrMU2 file
# key_new = directory + '/hyper_{1}_IS_{0}_{2}_points'.format(index_IS, new_func_name, num_pts)
# # print key_new
# send_data_to_s3(bucket, key_new, data)
# init data
num_replications = 100
num_pts = 10
for repl_no in xrange(num_replications):
print '\nrepl ' + str(repl_no)
for index_IS in func.getList_IS_to_query():
key = directory_backup + '/{2}_IS_{0}_{3}_points_repl_{1}'.format(
index_IS, repl_no, func_name, num_pts)
data = get_data_from_s3(bucket, key)
# write to lrMU2 file
key_new = directory + '/{2}_IS_{0}_{3}_points_repl_{1}'.format(
index_IS, repl_no, new_func_name, num_pts)
# print key_new
send_data_to_s3(bucket, key_new, data)
def hyper_param(self):
# with open("{0}/{1}_rb.pickle".format(self.hyper_dir, self.method_name), 'rb') as f:
# data = pickle.load(f)
data = get_data_from_s3(self._bucket, "coldstart/hyper/{0}_rb".format(self.method_name))
return data['hyperparam']
def hyper_param(self):
data = get_data_from_s3(
self._bucket, "coldstart/hyper/{0}_ato".format(self.method_name))
return data['hyperparam']
__author__ = 'matthiaspoloczek'
'''
Inspect the data used to generate hypers. The command line arg could be lrMU.
'''
conn = S3Connection()
conn = boto.connect_s3()
bucket = conn.get_bucket('poloczeks3', validate=True)
argv = sys.argv[1:]
func_name = argv[0]
directory = '/miso/data'
num_pts = 1000
key = directory + '/hyper_{1}_IS_{0}_{2}_points'.format(0, func_name, num_pts)
data0 = get_data_from_s3(bucket, key)
key = directory + '/hyper_{1}_IS_{0}_{2}_points'.format(1, func_name, num_pts)
data1 = get_data_from_s3(bucket, key)
### To inspect initial data
# num_pts = 10
# key = directory+'/{1}_IS_{0}_{2}_points_repl_0'.format(0, func_name, num_pts)
# data0 = get_data_from_s3(bucket, key)
# key = directory+'/{1}_IS_{0}_{2}_points_repl_0'.format(1, func_name, num_pts)
# data1 = get_data_from_s3(bucket, key)
print np.mean(data0["vals"])
print np.mean(data1["vals"])
#
# print np.mean(data0["noise"])
# print np.mean(data1["noise"])
Script to inspect hypers stored at S3
invoke as : python inspect_hypers.py miso_lrMU_hyper_ego
Optional parameters for lrMU:
"miso_lrMU_hyper_ego":
"miso_lrMU_hyper_mkg":
"miso_lrMU_hyper_pes":
"miso_lrMU_hyper_mei":
'''
conn = S3Connection()
conn = boto.connect_s3()
bucket = conn.get_bucket(s3_bucket_name, validate=True)
# construct problem instance given CMD args
argv = sys.argv[1:]
if argv[0].find("ego") < 0 and argv[0].find("kg") < 0 and argv[0].find("mei") < 0 and argv[0].find("mkg") < 0\
and argv[0].find("pes") < 0:
raise ValueError("No correct algo selected!")
problem = identify_problem(argv, bucket)
data = get_data_from_s3(bucket, problem.hyper_path)
print "prior_mean = " + str(data["prior_mean"])
print "prior_sig = " + str(data["prior_sig"])
if argv[0].find("pes") >=0:
print "hyperparam = " + str(data["hyperparam"])
print "hyperparam_mat = " + str(data["hyperparam_mat"])
else:
print "hyper_bounds = " + str(data["hyper_bounds"])
print "hyperparam = " + str(data["hyperparam"])
print "loglikelihood = " + str(data["loglikelihood"])
def hyper_param(self):
data = get_data_from_s3(
self._bucket, "miso/hyper/{0}_{1}".format(
self.method_name,
self._obj_func[self._obj_func_idx].getFuncName()))
return data['hyperparam']
__author__ = 'matthiaspoloczek'
'''
Script to inspect data stored at S3
invoke as : python inspect_results_s3.py miso_lrMU_benchmark_mkgcandpts 0
where 0 is a natural integer determining a replication
Optional parameters for lrMU:
"miso_lrMU_benchmark_ego":
"miso_lrMU_benchmark_mkg":
"miso_lrMU_benchmark_mkgcandpts":
"miso_lrMU_benchmark_pes":
"miso_lrMU_benchmark_mei":
'''
conn = S3Connection()
conn = boto.connect_s3()
bucket = conn.get_bucket(s3_bucket_name, validate=True)
# construct problem instance given CMD args
argv = sys.argv[1:]
if argv[0].find("ego") < 0 and argv[0].find("kg") < 0 and argv[0].find("mei") < 0 and argv[0].find("mkg") < 0\
and argv[0].find("pes") < 0:
raise ValueError("No correct algo selected!")
problem = identify_problem(argv, bucket)
data = get_data_from_s3(bucket, problem.result_path)
# print data['sampled_is']
print data['raw_voi']
