[REF] Improve numerical stability of Frobenius estimator (#120)

curvlinops/norm/hutchinson.py

@@ -1,8 +1,9 @@
 """Hutchinson-style matrix norm estimation."""
 
+from numpy import dot
 from scipy.sparse.linalg import LinearOperator
 
-from curvlinops.trace.hutchinson import HutchinsonTraceEstimator
+from curvlinops.sampling import random_vector
 
 
 class HutchinsonSquaredFrobeniusNormEstimator:
@@ -43,7 +44,7 @@ def __init__(self, A: LinearOperator):
         Args:
             A: Linear operator whose squared Frobenius norm will be estimated.
         """
-        self._trace_estimator = HutchinsonTraceEstimator(A.T @ A)
+        self._A = A
 
     def sample(self, distribution: str = "rademacher") -> float:
         """Draw a sample from the squared Frobenius norm estimator.
@@ -59,4 +60,7 @@ def sample(self, distribution: str = "rademacher") -> float:
         Returns:
             Sample from the squared Frobenius norm estimator.
         """
-        return self._trace_estimator.sample(distribution=distribution)
+        dim = self._A.shape[1]
+        v = random_vector(dim, distribution)
+        Av = self._A @ v
+        return dot(Av, Av)