Sped up nabla_f_w

shivamgupta2 · May 4, 2021 · 035feec · 035feec
1 parent 7758d10
commit 035feec
Showing 1 changed file with 13 additions and 1 deletion.
diff --git a/robustlib.py b/robustlib.py
@@ -11,6 +11,7 @@
 from matplotlib import rc
 import ast
 import mpld3
+import time
 #mpld3.enable_notebook()
 
 def err_rspca(a,b): return LA.norm(np.outer(a,a)-np.outer(b, b))
@@ -212,10 +213,15 @@ def alg(self, S, indicator):
         w = np.ones(m) / m
         X = S
         eps_m = round(eps * m)
+        nabla_f_w_total_time = 0
+        sigma_w_total_time = 0
+        start_time = time.time()
         for i in range(nItrs):
             if self.sparse:
                 Xw = X.T @ w
+                sigma_w_start_time = time.time()
                 Sigma_w = (X.T @ spdiags(w, 0, m, m) @ X) - (Xw @ Xw.T)
+                sigma_w_total_time += time.time() - sigma_w_start_time
                 Sigma_w_minus_I = Sigma_w - np.eye(d)
                 #find indices of largest k entries of each row of Sigma_w_minus_I
                 largest_k_each_row_index_array = np.argpartition(Sigma_w_minus_I, kth=-k, axis=-1)[:, -k:]
@@ -231,7 +237,10 @@ def alg(self, S, indicator):
                 psi_w[largest_rows_index_array, largest_k_each_row_index_array.T] = 1
                 Z_w = psi_w * Sigma_w_minus_I
 
-                nabla_f_w = ((X @ (Z_w @ X.T)).diagonal() - (X @ Z_w @ (X.T @ w)) - (X @ (Z_w.T @ (X.T @ w)))) / LA.norm(Z_w)
+                nabla_f_w_start_time = time.time()
+                nabla_f_w = ((X @ (Z_w @ X.T)).diagonal() - (X @ (Z_w @ (X.T @ w))) - (X @ (Z_w.T @ (X.T @ w)))) / LA.norm(Z_w)
+                #nabla_f_w = ((X @ (Z_w @ (X.T @ w))) - (X @ (Z_w.T @ (X.T @ w)))) / LA.norm(Z_w)
+                nabla_f_w_total_time += time.time() - nabla_f_w_start_time
             else:
                 Xw = np.matmul(X.T, w)
                 Sigma_w_fun = lambda v: np.matmul(X.T, w * np.matmul(X, v)) - Xw *np.matmul(Xw.T, v)
@@ -246,6 +255,9 @@ def alg(self, S, indicator):
             w = w - step_size * nabla_f_w/ LA.norm(nabla_f_w)
             w = self.project_onto_capped_simplex_simple(w, (1/(m - eps_m)))
             #print(w.shape)
+        print('Time to run GD ', time.time() - start_time)
+        print('Time to compute Sigma_w ', sigma_w_total_time)
+        print('Time to copute nabla_f_w', nabla_f_w_total_time)
         print(w.shape)
         print(X.shape)
         mu_gd = np.sum(w * X.T, axis=1)