Merge pull request #98 from teresamg/logosplit

ENH: Outsource leave-one-out splitter so it can be used across data types

src/eddymotion/data/dmri.py

@@ -71,64 +71,14 @@ class DWI:
     )
     """A path to an HDF5 file to store the whole dataset."""
 
+    def get_filename(self):
+        """Get the filepath of the HDF5 file."""
+        return self._filepath
+
     def __len__(self):
         """Obtain the number of high-*b* orientations."""
         return self.dataobj.shape[-1]
 
-    def logo_split(self, index, with_b0=False):
-        """
-        Produce one fold of LOGO (leave-one-gradient-out).
-
-        Parameters
-        ----------
-        index : :obj:`int`
-            Index of the DWI orientation to be left out in this fold.
-        with_b0 : :obj:`bool`
-            Insert the *b=0* reference at the beginning of the training dataset.
-
-        Returns
-        -------
-        (train_data, train_gradients) : :obj:`tuple`
-            Training DWI and corresponding gradients.
-            Training data/gradients come **from the updated dataset**.
-        (test_data, test_gradients) :obj:`tuple`
-            Test 3D map (one DWI orientation) and corresponding b-vector/value.
-            The test data/gradient come **from the original dataset**.
-
-        """
-        if not Path(self._filepath).exists():
-            self.to_filename(self._filepath)
-
-        # read original DWI data & b-vector
-        with h5py.File(self._filepath, "r") as in_file:
-            root = in_file["/0"]
-            dwframe = np.asanyarray(root["dataobj"][..., index])
-            bframe = np.asanyarray(root["gradients"][..., index])
-
-        # if the size of the mask does not match data, cache is stale
-        mask = np.zeros(len(self), dtype=bool)
-        mask[index] = True
-
-        train_data = self.dataobj[..., ~mask]
-        train_gradients = self.gradients[..., ~mask]
-
-        if with_b0:
-            train_data = np.concatenate(
-                (np.asanyarray(self.bzero)[..., np.newaxis], train_data),
-                axis=-1,
-            )
-            b0vec = np.zeros((4, 1))
-            b0vec[0, 0] = 1
-            train_gradients = np.concatenate(
-                (b0vec, train_gradients),
-                axis=-1,
-            )
-
-        return (
-            (train_data, train_gradients),
-            (dwframe, bframe),
-        )
-
     def set_transform(self, index, affine, order=3):
         """Set an affine, and update data object and gradients."""
         reference = namedtuple("ImageGrid", ("shape", "affine"))(

src/eddymotion/data/splitting.py

@@ -0,0 +1,84 @@
+# emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*-
+# vi: set ft=python sts=4 ts=4 sw=4 et:
+#
+# Copyright 2022 The NiPreps Developers <[email protected]>
+#
+# 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.
+#
+# We support and encourage derived works from this project, please read
+# about our expectations at
+#
+#     https://www.nipreps.org/community/licensing/
+#
+"""Data splitting helpers."""
+from pathlib import Path
+import numpy as np
+import h5py
+
+
+def lovo_split(dataset, index, with_b0=False):
+    """
+    Produce one fold of LOVO (leave-one-volume-out).
+
+    Parameters
+    ----------
+    dataset : :obj:`eddymotion.data.dmri.DWI`
+        DWI object
+    index : :obj:`int`
+        Index of the DWI orientation to be left out in this fold.
+
+    Returns
+    -------
+    (train_data, train_gradients) : :obj:`tuple`
+        Training DWI and corresponding gradients.
+        Training data/gradients come **from the updated dataset**.
+    (test_data, test_gradients) :obj:`tuple`
+        Test 3D map (one DWI orientation) and corresponding b-vector/value.
+        The test data/gradient come **from the original dataset**.
+
+    """
+
+    if not Path(dataset.get_filename()).exists():
+        dataset.to_filename(dataset.get_filename())
+
+    # read original DWI data & b-vector
+    with h5py.File(dataset.get_filename(), "r") as in_file:
+        root = in_file["/0"]
+        data = np.asanyarray(root["dataobj"])
+        gradients = np.asanyarray(root["gradients"])
+
+    # if the size of the mask does not match data, cache is stale
+    mask = np.zeros(data.shape[-1], dtype=bool)
+    mask[index] = True
+
+    train_data = data[..., ~mask]
+    train_gradients = gradients[..., ~mask]
+    test_data = data[..., mask]
+    test_gradients = gradients[..., mask]
+
+    if with_b0:
+        train_data = np.concatenate(
+            (np.asanyarray(dataset.bzero)[..., np.newaxis], train_data),
+            axis=-1,
+        )
+        b0vec = np.zeros((4, 1))
+        b0vec[0, 0] = 1
+        train_gradients = np.concatenate(
+            (b0vec, train_gradients),
+            axis=-1,
+        )
+
+    return (
+        (train_data, train_gradients),
+        (test_data, test_gradients),
+    )

src/eddymotion/estimator.py

@@ -34,6 +34,7 @@
 from pkg_resources import resource_filename as pkg_fn
 from tqdm import tqdm
 
+from eddymotion.data.splitting import lovo_split
 from eddymotion.model import ModelFactory
 
 
@@ -132,7 +133,7 @@ def fit(
                         pbar.set_description_str(
                             f"Pass {i_iter + 1}/{n_iter} | Fit and predict b-index <{i}>"
                         )
-                        data_train, data_test = dwdata.logo_split(i, with_b0=True)
+                        data_train, data_test = lovo_split(dwdata, i, with_b0=True)
                         grad_str = f"{i}, {data_test[1][:3]}, b={int(data_test[1][3])}"
                         pbar.set_description_str(f"[{grad_str}], {n_jobs} jobs")
 

test/test_model.py

@@ -26,6 +26,7 @@
 import pytest
 
 from eddymotion import model
+from eddymotion.data.splitting import lovo_split
 from eddymotion.data.dmri import DWI
 
 
@@ -97,7 +98,7 @@ def test_two_initialisations(datadir):
     dmri_dataset = DWI.from_filename(datadir / "dwi.h5")
 
     # Split data into test and train set
-    data_train, data_test = dmri_dataset.logo_split(10)
+    data_train, data_test = lovo_split(dmri_dataset, 10)
 
     # Direct initialisation
     model1 = model.AverageDWModel(

test/test_splitting.py

@@ -0,0 +1,62 @@
+# emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*-
+# vi: set ft=python sts=4 ts=4 sw=4 et:
+#
+# Copyright 2021 The NiPreps Developers <[email protected]>
+#
+# 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.
+#
+# We support and encourage derived works from this project, please read
+# about our expectations at
+#
+#     https://www.nipreps.org/community/licensing/
+#
+"""Unit test testing the lovo_split function."""
+import numpy as np
+from eddymotion.data.dmri import DWI
+from eddymotion.data.splitting import lovo_split
+
+
+def test_lovo_split(datadir):
+    """
+    Test the lovo_split function.
+
+    Parameters:
+    - datadir: The directory containing the test data.
+
+    Returns:
+    None
+    """
+    data = DWI.from_filename(datadir / "dwi.h5")
+
+    # Set zeros in dataobj and gradients of the dwi object
+    data.dataobj[:] = 0
+    data.gradients[:] = 0
+
+    # Select a random index
+    index = np.random.randint(len(data))
+
+    # Set 1 in dataobj and gradients of the dwi object at this specific index
+    data.dataobj[..., index] = 1
+    data.gradients[..., index] = 1
+
+    # Apply the lovo_split function at the specified index
+    (train_data, train_gradients), \
+        (test_data, test_gradients) = lovo_split(data, index)
+
+    # Check if the test data contains only 1s
+    # and the train data contains only 0s after the split
+    assert np.all(test_data == 1)
+    assert np.all(train_data == 0)
+    assert np.all(test_gradients == 1)
+    assert np.all(train_gradients == 0)
+