SIT

python/ad-hoc/SIT.py

@@ -0,0 +1,19 @@
+import  numpy as np
+def SIT(A, b):
+  # if the matrix is not lower triangular, we cannot solve the system
+  if not np.allclose(A, np.tril(A)):
+    print("The matrix A is not lower triangular!")
+    return np.nan
+
+  n = len(b)
+  x = np.zeros((n, 1))
+
+  # calculate x1
+  x[0, 0] = b[0, 0] / A[0, 0]
+
+  # calculate x(i) forwards
+  for i in range (1, n):
+    sum_of_xs = np.dot(A[i, 0 : i], x[0 : i, 0])
+    x[i, 0] = (b[i, 0] - sum_of_xs) / A[i, i]
+
+  return x