Random feature layout in simulator

.gitignore

@@ -174,4 +174,5 @@ Thumbs.db
 
 src/cpp/third_party
 test_datasets
-simulator/datasets
+simulator/datasets
+simulator/images

simulator/configs/arvix.json

@@ -1,6 +1,7 @@
 {
     "dataset_name" : "ogbn_arxiv",
     "features_stats" : {
+        "feature_layout" : "random",
         "page_size" : "16 KB",
         "feature_dimension" : 128,
         "feature_size" : "float32"

simulator/main.py

@@ -16,13 +16,12 @@ def read_config_file(config_file):
 
 def read_arguments():
     parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter)
-    parser.add_argument("config_file", type=str, help="The config file containing the details for the simulation")
+    parser.add_argument("--config_file", type=str, help="The config file containing the details for the simulation")
+    parser.add_argument("--save_path", required=True, type=str, help="The path to save the resulting image to")
+    parser.add_argument("--graph_title", required=True, type=str, help="The title of the saved graph")
     return parser.parse_args()
 
 
-IMAGES_SAVE_DIR = "images"
-
-
 def main():
     arguments = read_arguments()
     config = read_config_file(arguments.config_file)
@@ -44,8 +43,8 @@ def main():
     print("Got result for", len(pages_loaded), "nodes out of", len(nodes_to_sample), "nodes")
 
     # Save the histogram
-    save_path = os.path.join(IMAGES_SAVE_DIR, os.path.basename(arguments.config_file).replace("json", "png"))
-    visualize_results(pages_loaded, save_path, config["dataset_name"])
+    os.makedirs(os.path.dirname(arguments.save_path), exist_ok=True)
+    visualize_results(pages_loaded, arguments.save_path, arguments.graph_title, config["dataset_name"])
 
 
 if __name__ == "__main__":

simulator/src/features_loader.py

@@ -1,6 +1,7 @@
 import humanfriendly
 import os
 import math
+import random
 
 
 class FeaturesLoader:
@@ -11,9 +12,16 @@ def __init__(self, data_loader, features_stat):
         self.node_feature_size = self.feature_size * features_stat["feature_dimension"]
 
         self.nodes_per_page = max(int(self.page_size / self.node_feature_size), 1)
-        self.total_pages = int(math.ceil(data_loader.get_num_nodes() / (1.0 * self.nodes_per_page)))
+        total_nodes = data_loader.get_num_nodes()
+        self.total_pages = int(math.ceil(total_nodes / (1.0 * self.nodes_per_page)))
+
+        self.node_location_map = [i for i in range(total_nodes)]
+        if "feature_layout" in features_stat and features_stat["feature_layout"] == "random":
+            random.shuffle(self.node_location_map)
+            print(self.node_location_map[:10])
 
     def get_node_page(self, node_id):
+        node_location = self.node_location_map[node_id]
         return int(node_id / self.nodes_per_page)
 
     def get_total_file_size(self):

simulator/src/visualizer.py

@@ -2,14 +2,14 @@
 import os
 
 
-def visualize_results(pages_loaded, save_path, dataset_name, num_bins=50):
+def visualize_results(pages_loaded, save_path, graph_title, dataset_name, num_bins=50):
     # Create the histogram
     plt.figure()
     plt.ecdf(pages_loaded, label="CDF")
     plt.hist(pages_loaded, bins=num_bins, histtype="step", density=True, cumulative=True, label="Cumulative histogram")
     plt.xlabel("Number of pages loaded for node inference")
     plt.ylabel("Percentage of nodes")
-    plt.title("Number of pages loaded for node inference on " + dataset_name)
+    plt.title(graph_title)
     plt.xlim(0, 50)
     plt.legend()