test_compute_frd_metric.py

# Copyright (c) MONAI Consortium
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#     http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import annotations

import unittest

import numpy as np
import torch

from monai.metrics import FIDMetric, FrechetRadiomicsDistance
from monai.utils import optional_import

_, has_scipy = optional_import("scipy")


@unittest.skipUnless(has_scipy, "Requires scipy")
class TestFrechetRadiomicsDistance(unittest.TestCase):
    def test_results(self):
        x = torch.Tensor([[1, 2], [1, 2], [1, 2]])
        y = torch.Tensor([[2, 2], [1, 2], [1, 2]])
        results = FrechetRadiomicsDistance()(x, y)
        np.testing.assert_allclose(results.cpu().numpy(), 0.4444, atol=1e-4)

    def test_frd_matches_fid_for_same_features(self):
        """FRD uses the same Fréchet formula as FID; same inputs give same value."""
        y_pred = torch.Tensor([[1.0, 2.0], [1.0, 2.0], [1.0, 2.0]])
        y = torch.Tensor([[2.0, 2.0], [1.0, 2.0], [1.0, 2.0]])
        frd_score = FrechetRadiomicsDistance()(y_pred, y)
        fid_score = FIDMetric()(y_pred, y)
        np.testing.assert_allclose(frd_score.cpu().numpy(), fid_score.cpu().numpy(), atol=1e-6)

    def test_rejects_high_dimensional_input(self):
        """FrechetRadiomicsDistance raises ValueError when inputs have ndimension() > 2."""
        high_dim = torch.ones([3, 3, 144, 144])
        with self.assertRaises(ValueError):
            FrechetRadiomicsDistance()(high_dim, high_dim)


if __name__ == "__main__":
    unittest.main()