def learning_pipeline(data, conf, config_path='snapw.config'):
master_server, server_list = get_servers()
data_c = threading.Thread(target=poll_for_kv,
args=(data, server_list, master_server))
data_c.start()
while data_c.isAlive():
time.sleep(1)
logging.info("Learning data captured.")
# Conf is a dictionary with key as paramter, value as value, i.e. k-v file.
# Data is dictionary with key as machine, value as the k-k-v file.
# Setting is a dictionary with key as parameter, value as the category.
logging.critical("DATA: %s", data)
label_rt = 1.0
setting = {
"GenTasks": "average",
"GenStubs": "average",
"GenGraph": "average",
"GetNbr": "average"
}
logging.info("Generating features.")
features, target = generate_features(label_rt, conf["var"], data, setting)
logging.info("Training model.")
train(features, target)
logging.info("Generating new configuration file")
new_conf_d = predict(conf["var"])
return new_conf_d
def learning_pipeline(data, conf, config_path='snapw.config'):
master_server, server_list = get_servers()
data_c = threading.Thread(target=poll_for_kv, args=(data, server_list, master_server ))
data_c.start()
while data_c.isAlive():
time.sleep(1)
logging.info("Learning data captured.")
# Conf is a dictionary with key as paramter, value as value, i.e. k-v file.
# Data is dictionary with key as machine, value as the k-k-v file.
# Setting is a dictionary with key as parameter, value as the category.
logging.critical("DATA: %s", data)
label_rt = 1.0
setting = {"GenTasks": "average", "GenStubs":"average", "GenGraph":"average", "GetNbr":"average"}
logging.info("Generating features.")
features, target = generate_features(label_rt, conf["var"], data, setting)
logging.info("Training model.")
train(features, target)
logging.info("Generating new configuration file")
new_conf_d = predict(conf["var"])
return new_conf_d
# Conf is a dictionary with key as paramter, value as value, i.e. k-v file,
# Data is dictionary with key as machine, value as the k-k-v file,
# setting is a dictionary with key as parameter, value as the category,
# features, target = generate_features(conf, data, setting)
# train(features, target)
# new_conf_d = predict(features)
# Generate new configuration file and pass to other machines.
label = 1.0
conf = {"nodes":10000, "range":1, "stat_tasks":1, "gen_tasks":10000, "drange":10000}
data = {"GenTasks":{1:2.0, 2:3.0, 4:2.4}, "GenStubs":{"a":1.0, "b":2.0, "c":3}, "GenGraph":{"1":101.0, "2":99.0, "4":2.4}, "GetNbr":{"1":1.0, "2":9.0, "4":1.4}, "GetDist":{"1":101.0, "2":99.0, "3":2.4}}
setting = {"GenTasks": "average", "GenStubs":"average", "GenGraph":"average", "GetNbr":"average"}
features, target = generate_features(label, conf, data, setting)
train(features, target)
new_conf_d = predict(conf)
print new_conf_d
'''
route __Start__ GenTasks
route GenTasks GenStubs
route GenStubs GenGraph
route GenGraph GetNbr
route GetNbr:1 GetDist
route GetNbr:2 GetTargets
route GetTargets GetDist
route GetDist:1 GetNbr
route GetDist:2 __Finish__
'''
# new_conf_d = predict(features)
# Generate new configuration file and pass to other machines.
label = 1.0
conf = {"nodes": 10000, "range": 1, "stat_tasks": 1, "gen_tasks": 10000, "drange": 10000}
data = {
"GenTasks": {1: 2.0, 2: 3.0, 4: 2.4},
"GenStubs": {"a": 1.0, "b": 2.0, "c": 3},
"GenGraph": {"1": 101.0, "2": 99.0, "4": 2.4},
"GetNbr": {"1": 1.0, "2": 9.0, "4": 1.4},
"GetDist": {"1": 101.0, "2": 99.0, "3": 2.4},
}
setting = {"GenTasks": "average", "GenStubs": "average", "GenGraph": "average", "GetNbr": "average"}
features, target = generate_features(label, conf, data, setting)
train(features, target)
new_conf_d = predict(conf)
print new_conf_d
"""
route __Start__ GenTasks
route GenTasks GenStubs
route GenStubs GenGraph
route GenGraph GetNbr
route GetNbr:1 GetDist
route GetNbr:2 GetTargets
route GetTargets GetDist
route GetDist:1 GetNbr
route GetDist:2 __Finish__
"""