Merge pull request #217 from EmmaRenauld/follow_official_scilpy

Follow official scilpy

.github/workflows/test_package.yml

@@ -27,7 +27,6 @@ jobs:
       run: |
         pip install --upgrade pip
         pip install pytest
-        pip install -r requirements_github.txt
         pip install -e .
 
     - name: Tests

dwi_ml/data/processing/space/neighborhood.py

@@ -1,6 +1,5 @@
 # -*- coding: utf-8 -*-
 import itertools
-from typing import Union, List
 
 import numpy as np
 import torch

dwi_ml/data/processing/streamlines/data_augmentation.py

@@ -7,7 +7,8 @@
 from dipy.io.stateful_tractogram import StatefulTractogram
 from nibabel.streamlines.tractogram import (PerArrayDict, PerArraySequenceDict)
 import numpy as np
-from scilpy.tractograms.streamline_operations import resample_streamlines_step_size
+
+from scilpy.tracking.tools import resample_streamlines_step_size
 from scilpy.utils.streamlines import compress_sft
 
 

dwi_ml/data/processing/streamlines/post_processing.py

@@ -1,13 +1,12 @@
 # -*- coding: utf-8 -*-
-import logging
 from typing import List
 
 import numpy as np
 import torch
 
-from scilpy.tractograms.uncompress import uncompress
 from scilpy.tractanalysis.tools import \
     extract_longest_segments_from_profile as segmenting_func
+from scilpy.tractanalysis.uncompress import uncompress
 
 # We could try using nan instead of zeros for non-existing previous dirs...
 DEFAULT_UNEXISTING_VAL = torch.zeros((1, 3), dtype=torch.float32)

dwi_ml/models/embeddings.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-from typing import Tuple, Union, List
+from typing import Tuple, List
 
 import numpy as np
 import torch

dwi_ml/models/projects/learn2track_model.py

@@ -11,10 +11,10 @@
     compute_directions, normalize_directions, compute_n_previous_dirs
 from dwi_ml.data.processing.streamlines.sos_eos_management import \
     convert_dirs_to_class
-from dwi_ml.models.embeddings import NoEmbedding, keys_to_embeddings
+from dwi_ml.models.embeddings import NoEmbedding
 from dwi_ml.models.main_models import (
     ModelWithPreviousDirections, ModelWithDirectionGetter,
-    ModelWithNeighborhood, MainModelOneInput, ModelOneInputWithEmbedding)
+    ModelWithNeighborhood, ModelOneInputWithEmbedding)
 from dwi_ml.models.stacked_rnn import StackedRNN
 
 logger = logging.getLogger('model_logger')  # Same logger as Super.

dwi_ml/models/projects/learn2track_utils.py

@@ -1,7 +1,6 @@
 # -*- coding: utf-8 -*-
 import argparse
 
-from dwi_ml.models.embeddings import keys_to_embeddings
 from dwi_ml.models.projects.learn2track_model import Learn2TrackModel
 
 

dwi_ml/models/projects/transformers_utils.py

@@ -1,10 +1,8 @@
 # -*- coding: utf-8 -*-
-from dwi_ml.models.embeddings import keys_to_embeddings
 from dwi_ml.models.positional_encoding import (
     keys_to_positional_encodings)
 from dwi_ml.models.projects.transformer_models import (
     AbstractTransformerModel)
-from dwi_ml.models.utils.direction_getters import check_args_direction_getter
 
 sphere_choices = ['symmetric362', 'symmetric642', 'symmetric724',
                   'repulsion724', 'repulsion100', 'repulsion200']

dwi_ml/testing/projects/transformer_visualisation_utils.py

@@ -10,7 +10,7 @@
 
 from scilpy.io.streamlines import load_tractogram_with_reference
 from scilpy.io.utils import add_reference_arg, add_overwrite_arg, add_bbox_arg
-from scilpy.tractograms.streamline_operations import resample_streamlines_step_size
+from scilpy.tracking.tools import resample_streamlines_step_size
 from scilpy.utils.streamlines import compress_sft
 
 from dwi_ml.io_utils import add_logging_arg

dwi_ml/training/batch_samplers.py

@@ -28,8 +28,6 @@
 from typing import List, Tuple, Iterator, Union
 
 import numpy as np
-import torch
-import torch.multiprocessing
 from torch.utils.data import Sampler
 
 from dwi_ml.data.dataset.multi_subject_containers import MultiSubjectDataset

dwi_ml/training/trainers.py

@@ -4,7 +4,7 @@
 import logging
 import os
 import shutil
-from typing import Union, List, Tuple
+from typing import Union, List
 
 from comet_ml import (Experiment as CometExperiment, ExistingExperiment)
 import numpy as np

dwi_ml/training/utils/batch_samplers.py

@@ -1,8 +1,6 @@
 # -*- coding: utf-8 -*-
 import argparse
-import logging
 
-from dwi_ml.experiment_utils.prints import format_dict_to_str
 from dwi_ml.experiment_utils.timer import Timer
 from dwi_ml.training.batch_samplers import DWIMLBatchIDSampler
 

dwi_ml/training/with_generation/batch_loader.py

@@ -1,7 +1,5 @@
 # -*- coding: utf-8 -*-
-from typing import List, Dict
-
-import torch
+from typing import Dict
 
 from dwi_ml.training.batch_loaders import DWIMLBatchLoaderOneInput
 

dwi_ml/unit_tests/test_dataset.py

@@ -4,7 +4,6 @@
 import os
 
 import h5py
-import numpy as np
 import torch
 
 from dwi_ml.data.dataset.multi_subject_containers import \

dwi_ml/unit_tests/test_submethods_interpolation.py

@@ -5,8 +5,7 @@
 import torch
 
 from dwi_ml.data.processing.space.neighborhood import \
-    prepare_neighborhood_vectors, extend_coordinates_with_neighborhood, \
-    unflatten_neighborhood
+    prepare_neighborhood_vectors, unflatten_neighborhood
 from dwi_ml.data.processing.volume.interpolation import \
     interpolate_volume_in_neighborhood
 

requirements.txt

@@ -1,12 +1,9 @@
 # Supported for python 3.10
 # Should work for python > 3.8.
 
-# ----------
-# Scilpy and dipy must be installed manually first.
-# (or run pip install -r requirements_github.txt)
-# Not adding here to let you use your favorite version.
-# ----------
-
+# Scilpy and comet_ml both require requests. In comet: >=2.18.*,
+# which installs a version >2.28. Adding request version explicitely.
+requests==2.28.*
 bertviz~=1.4.0  # For transformer's visu
 torch==1.13.*
 tqdm==4.64.*
@@ -16,7 +13,7 @@ jupyterlab>=3.6.2  # For transformer's visu
 IProgress>=0.4     # For jupyter with tdqm
 nested_lookup==0.2.25
 nose==1.3.*
-
+scilpy==1.5.post2
 
 ## Necessary but should be installed with scilpy (Last check: 09/2023):
 future==0.18.*
@@ -27,10 +24,6 @@ numpy==1.23.*
 scipy==1.9.*
 
 
-# Scilpy requires requests==2.28.*, but comet_ml requires requests>=2.18.*,
-# which installs a version >2.28. Adding request version explicitely.
-requests==2.28.*
-
 
 # --------------- Notes to developers
 # If we upgrade torch, verify if code copied in

scripts_python/tests/test_all_steps_tts.py

@@ -5,11 +5,8 @@
 import pytest
 import tempfile
 
-import torch
-
 from dwi_ml.unit_tests.utils.expected_values import \
-    (TEST_EXPECTED_VOLUME_GROUPS, TEST_EXPECTED_STREAMLINE_GROUPS,
-     TEST_EXPECTED_SUBJ_NAMES)
+    (TEST_EXPECTED_VOLUME_GROUPS, TEST_EXPECTED_STREAMLINE_GROUPS)
 from dwi_ml.unit_tests.utils.data_and_models_for_tests import fetch_testing_data
 
 data_dir = fetch_testing_data()