make change so code be run more easily

README.md

@@ -1,30 +1,12 @@
-The "src" folder is the source code of SAMI paper which has seen some minor changes 
+The "MissingNodes_SrcCode2014" folder is the source code of SAMI paper which has seen some minor changes 
 from the original received version.
-The "codes" folder is the folder containing the following files: 
-1.adjlisttext_tonxgraph.py: gets an adj. list as input file and gives out the visualized
- graph
-2.mat_tonxgraph.py: gets a .mat file as input file and gives out the visualized graph
-3.mae.py: computes mae of two graphs given their .mat adj matrix file
 
-For using the source code you must pay attention to two files: "RunExpSrv8s" and 
-"MissingNodes_S8b". The first one is the function which sets up the files and 
-calls the main function i.e. "MissingNodes_S8b".
+**Running the project**
+For running the source code you must use a command like "RunExpSrv8s('0','0','1')" where first parameter shows start of iteration, 
+second parameter shows end of iteration and last param is an arbitrary number.
 
+**Input**
+You must place all input files in a folder named "in" in root directory of project.
 
-For having your own dataset run on this code you should know the below variables:
-
-*RUNEXPSRV_8S.m
-1. num_missing_nodes_arr: number of randomly removed nodes on each iteration
-2. ds_*: array of node numbers of your input arr
-3. ds_str: dataset path as a string
-4. root_dir
-5. file_prefix: file name prefix of the input files
-
-*MISSINGNODES_S8B
-1. k: filename postfix for saving
-2. clusters_out: true and predicted graph saved as .mat file
-3. graphs_out: 4 graphs including {original, predicted, predicted permuted, 
-original permuted} as .mat file
-
-
-*grid dataset is run on the src code as an example
+**Calcuating MAE**
+For calculating the MEA you must first run the project to get the output then run the python code named "calc_mae.py"

calc_mae.py

@@ -0,0 +1,17 @@
+import numpy as np
+import networkx as nx
+import matplotlib.pyplot as plt
+import scipy.io as sp
+import os
+
+path = 'output\mine'
+missing_node_count = 2
+mae_list = list()
+for filename in os.listdir(path):
+    adj_matrix_orig = sp.loadmat(path + '\\' + filename)['remapped_data'].A
+    adj_matrix_p = sp.loadmat(path + '\\' + filename)['out_data'].A
+    mae = np.sum(abs(adj_matrix_orig - adj_matrix_p))
+    mae = mae / (adj_matrix_p.shape[0] * 2 * missing_node_count)
+    mae_list.append(mae)
+
+print(np.mean(mae_list))