import argparse, os
parser = argparse.ArgumentParser(
description=
'Runs the approximated self similarity join (1NN) algorithm of a given ser of points several times'
)
parser.add_argument('input_matrix',
type=str,
help='The numpy vector storing file')
parser.add_argument('output_folder',
type=str,
help='The directory where the results must be stored')
args = parser.parse_args()
data = []
for line in open(args.input_matrix, "r"):
data.append(line.strip())
k = 1
for c in [1, 2, 3]:
for i in range(100):
print("running %d experiment" % i)
results = knn.sim_join(data, k, c)
f = open(
os.path.join(args.output_folder, str(k), str(c), "%d.res" % i),
"w")
for x, nn in results:
f.write("%d,%s\n" % (x, str(nn)))
f.close()
)
parser.add_argument('input_matrix',
type=str,
help='The numpy vector storing file')
parser.add_argument('output_folder',
type=str,
help='The directory where the results must be stored')
parser.add_argument(
'--N',
dest='iter',
type=int,
default=1,
help='The number of times the experiment must be repeated. 1 by default.')
parser.add_argument(
'--k',
dest='k',
type=int,
default=10,
help='The number of nearest neighbors to retrieve. 10 by default.')
args = parser.parse_args()
data = np.load(args.input_matrix)
for i in range(args.iter):
print("running %d experiment" % i)
results = knn.sim_join(data, args.k, 2)
f = open(os.path.join(args.output_folder, "%d.res" % i), "w")
for x, nn in results:
f.write("%d,%s\n" % (x, str(nn)))
f.close()