Sp2

data-generation/datagen.py

@@ -1,66 +1,43 @@
+import json
 from random import randint
-import csv
-import sys
 
 #-------------------------------------------------------------------------
-# Generates data in 3d space 0 to 1000 cube.
-# Outputs a csv file in format cluster number, x, y, z
-# Parameters:
-#   cluster_size integer - number of points to generate in each cluster
-#   cluster_number integer - number of clusters in space, capped at 8
-#   file_name string - name of file to save the cluster
-#   noise boolean - True - generate only clusters, False - generate some more random points
-def generate_3d_data(file_name, cluster_number, cluster_size, noise):
-
-    if (cluster_number) > 8:
-        return False
-
-    with open(file_name, 'w', newline='') as csv_file:
-        data_writer = csv.writer(csv_file, delimiter=',', quotechar='|', quoting=csv.QUOTE_MINIMAL)
-        for cluster in range(0, cluster_number):
-            x = 100 + 500 * ((cluster >> 2) % 2)
-            y = 100 + 500 * ((cluster >> 1) % 2)
-            z = 100 + 500 * ((cluster >> 0) % 2)
-            for item in range(1, cluster_size + 1):
-                data_writer.writerow([cluster, randint(0, 450) + x, randint(0, 450)
-                     + y, randint(0, 450) + z])
-    if noise == True:
-        generate_random_points(cluster_size, file_name)
-    return True
-
-#-------------------------------------------------------------------------
-# Generates random points in 0 to 1000 space and appends them to csv file
-# Format 100, x, y, z where 100 stands for no cluster but scattered random points
-# Parameters:
-#   number_of_points integer - how many points to generate
-#   file_name string - to what file to append
-def generate_random_points(number_of_points, file_name):
-    with open(file_name, 'a', newline='') as csv_file:
-        data_writer = csv.writer(csv_file, delimiter=',', quotechar='|', quoting=csv.QUOTE_MINIMAL)
-        for i in range(0, number_of_points + 1):
-            data_writer.writerow([100, randint(0, 1000), randint(0, 1000), randint(0, 1000)])
+# filling table = for size = 4
+#  XX | a0 | a1 | a2 | a3
+# ------------------------
+#  a0 | XX |    |    |
+# ------------------------
+#  a1 |fill| XX |    |
+# ------------------------
+#  a2 |fill|fill| XX |
+# ------------------------
+#  a3 |fill|fill|fill| XX
+# ------------------------
+def generate_similarity_tables(size):
+
+    ret_dict = {}
+    id_num = 0
+    while id_num != size: #generate identificators
+        ret_dict.update({"a" + str(id_num) : []})
+        id_num += 1
+
+    for row in ret_dict:
+        for column in ret_dict:
+            if(row == column):
+                continue
+            app = randint(0,1000)
+            if (app > 800):
+                ret_dict[row].append({column:app})
+                ret_dict[column].append({row:app})
+    return ret_dict
 
 #-------------------------------------------------------------------------
 
-print("Data generation started:")
+print("Starting to generate data")
 
-if len(sys.argv) != 5:
-    print("Usage: python3 datagen.py <output filename> <number of clusters> <points per cluster> <generate noise>")
-    exit(1)
+data = generate_similarity_tables(1000)
 
-out_file = sys.argv[1]
-number_of_clusters = int(sys.argv[2])
-points_per_cluster = int(sys.argv[3])
-if (sys.argv[4] == 'T'):
-    gen_noise = True
-elif (sys.argv[4] == 'F'):
-    gen_noise = False
-else:
-    print("Please use T (true) or F (false) to generate specify whether you want to generate noise.")
-    exit(1)
 
-if generate_3d_data(out_file, number_of_clusters, points_per_cluster, gen_noise):
-    print("Data generated.")
-else:
-    print("Data generation failed.")
+output_file = open("data.json", 'w')
 
+output_file.write(json.dumps(data))

datasets/overlap.csv

@@ -2997,4 +2997,4 @@
 5,607,375,477
 5,489,532,480
 5,763,350,645
-5,357,510,698
+5,357,510,698