Fix lanczos test.

Matrix was badly conditioned.

README.md

@@ -4,7 +4,7 @@ Project description
 
 ---
 
-## 🚀 Quick Start
+## Quick Start
 
 ### Setup
 
@@ -19,3 +19,13 @@ Project description
 - **Test**: `uv run pytest`
 - **Format**: `uv run ruff format .`
 - **Lint**: `uv run ruff check . --fix`
+
+## References
+
+[1] Susnjara et al., Accelerated filtering on graphs using Lanczos method.
+
+[2] https://epfl-lts2.github.io/gspbox-html/
+
+## Online guides
+
+https://every-algorithm.github.io/2024/05/23/lanczos_algorithm.html

src/afgl/lanczos.py

@@ -1,6 +1,12 @@
 import numpy as np
+import numpy.linalg as LA
 
 """
+Arguments
+L Real valued NxN symmetric matrix
+s vector of size N
+M natural number indicating basis size
+
 Returns
 -------
 V : ndarray
@@ -13,16 +19,22 @@ beta : ndarray
 
 
 def lanczos(L, s, M):
+    N = len(s)
     alp = np.zeros(M)
-    beta = np.zeros(M)
-    V = np.zeros(M)
-    V[0] = s / np.norm(s)
-
-    for j in range(0, M):
-        w = L * V
-        alp[j] = V[j] @ w
-        Vtmp = w - V[j] * alp[j]
-        if j > 1:
-            Vtmp = Vtmp - V[j - 1] * beta[j - 1]
-        beta[j] = np.norm(Vtmp)
-        V[j + 1] = Vtmp / beta[j]
+    beta = np.zeros(M - 1)
+    V = np.zeros((N, M))
+    V[:, 0] = s / LA.norm(s)
+
+    for j in range(M):
+        w = L @ V[:, j]
+        alp[j] = np.dot(V[:, j], w)
+
+        v_tilde = w - V[:, j] * alp[j]
+        if j > 0:
+            v_tilde = v_tilde - V[:, j - 1] * beta[j - 1]
+
+        if j < M - 1:
+            beta[j] = LA.norm(v_tilde)
+            V[:, j + 1] = v_tilde / beta[j]
+
+    return [V, alp, beta]

tests/test_main.py

@@ -0,0 +1,29 @@
+import numpy as np
+import numpy.linalg as LA
+from afgl.lanczos import lanczos
+
+"""
+Todo: better test case
+"""
+
+
+def test_lanczos_return_correct_solution():
+    N = 1000
+    M = 999
+
+    eigvals = np.random.uniform(10000, 100000, N)
+    Q, _ = LA.qr(np.random.randn(N, N))
+    L = Q @ np.diag(eigvals) @ Q.T
+
+    s = np.random.randint(1, 10, N)
+    [V, alp, beta] = lanczos(L, s, M)
+
+    T = np.diag(alp) + np.diag(beta, -1) + np.diag(beta, 1)
+
+    x = LA.solve(L, s)
+    e_1 = np.zeros(M)
+    e_1[0] = 1
+    y = (LA.inv(T) @ e_1) * LA.norm(s)
+    x_lanczos = V @ y
+
+    assert LA.norm(x - x_lanczos) < 1e-10