def main():
# file name will start with "als_prefix", which looks like "als3_"
file_name_suffix = "eval_results.bin"
# keep track of average and number of data points for each test
average_agreement = []
num_data_points = []
factor_list = [3, 5, 7, 9, 11]
for factor in factor_list:
als_prefix = "als" + str(factor) + "_"
# issue "als_eval" test for the current factor and print results
length = movie_lens_data.get_als_obj(als_prefix +
"user_ratings_test_length")
# sends an "als_eval" command to cluster nodes.
if cluster.cluster_info is None:
print("Cannot connect to cluster.")
return
cluster.send_command({
"op": "distribute",
"worker_op": "als_eval",
"setup_param": factor,
"length": length
})
cluster.wait_for_completion()
cluster.print_status()
print()
results = cluster.merge_list_results(als_prefix + file_name_suffix)
user_ids, agreements = zip(*results)
my_util.print_rank_agreement_results(agreements, "ALS " + str(factor))
print()
# update "average_agreement" and "num_data_points"
average_agreement.append(sum(agreements) / len(agreements))
num_data_points.append(len(agreements))
# print a summary table for all factors
print("factor".center(10), "average agreement".center(25),
"number of values".center(20))
for i in range(0, len(factor_list)):
print(
str(factor_list[i]).center(10),
str(average_agreement[i]).center(25),
str(num_data_points[i]).center(20))
print('\a')
def _register_with_cluster(self):
"""Register with cluster server. Return True for success."""
try:
reply = cluster.send_command(
{
"op": "register",
"address": self.address,
"port_number": port_number,
"cpu_count": multiprocessing.cpu_count()
},
wait_for_reply=True)
if reply["op"] == "reply":
# print cluster information
cluster_info = cluster.cluster_info
print("Worker server at", self.address, "port number",
port_number)
print("Joined cluster at", cluster_info[0]["address"],
"port number", cluster_info[0]["port_number"])
return True
else:
return False
except Exception as ex:
print("Error:", ex)
return False
def build_with_cluster(buff_point, buff_limit, length, output_file_name):
setup_param = {"buff_point": buff_point, "buff_limit": buff_limit}
cluster.send_command({
"op": "distribute",
"worker_op": "build_similar_movies_db",
"setup_param": setup_param,
"length": length
})
cluster.wait_for_completion()
cluster.print_status()
print()
cluster.merge_list_results_into_dict(output_file_name)
print("similar_movies.bin has been built")
print('\a')
def main():
print("Creating input.bin")
x1, x2 = _setup_problem(10000)
# Trigger an add operation by sending a distributed
# add command to cluster server.
if cluster.cluster_info is None:
print("Cannot connect to cluster.")
return
cluster.send_command({
"op": "distribute",
"worker_op": "add",
"length": len(x1)
})
cluster.wait_for_completion()
cluster.print_status()
cluster.merge_list_results("output.bin")
_check_output_bin()
def main():
length = movie_lens_data.get_input_obj("user_ratings_test_length")
# Sends a "tag_ls_eval" command to cluster nodes.
if cluster.cluster_info is None:
print("Cannot connect to cluster.")
return
cluster.send_command({
"op": "distribute",
"worker_op": "tag_ls_eval",
"length": length
})
cluster.wait_for_completion()
cluster.print_status()
print()
file_name = "tag_ls_eval_results.bin"
results = cluster.merge_list_results(file_name)
user_ids, agreements = zip(*results)
my_util.print_rank_agreement_results(agreements, "tag least squares")
print('\a')
def _send_done_reply(self):
cluster.send_command({
"op": "done",
"address": self.address,
"port_number": port_number
})