Merge branch '192-model_zoo_tests' into 'dev'

"Create tests for demos and model zoo entries"

Closes NifTK#192

See merge request CMIC/NiftyNet!227

.gitlab-ci.yml

@@ -29,6 +29,7 @@ testjob:
     - dev
     - tags
     - 195-add-evaluation-action-3
+    - 192-model_zoo_tests
   script:
     # !!kill coverage in case of hanging processes
     - if pgrep coverage; then pkill -f coverage; fi
@@ -120,6 +121,7 @@ testjob:
     - python net_autoencoder.py inference -c config/vae_config.ini --inference_type encode --save_seg_dir output/vae_demo_features
     - python net_autoencoder.py inference -c config/vae_config.ini --inference_type encode-decode
 
+    - python -m tests.test_model_zoo
     - python -m unittest discover -s "tests" -p "*_test.py"
 
     # deactivate virtual environment
@@ -174,6 +176,7 @@ testjob:
 
     - coverage run -a --source . net_download.py dense_vnet_abdominal_ct_model_zoo -r
 
+    - coverage run -a --source . -m tests.test_model_zoo
     - coverage run -a --source . -m unittest discover -s "tests" -p "*_test.py"
     - coverage report -m
     - echo 'finished test'
@@ -200,6 +203,7 @@ quicktest:
     - 221-add-changelog-entry-for-version-0.2.2
     - 223-put-bug-fixes-under-fixed-header-in-changelog
     - 195-add-evaluation-action-3
+    - 192-model_zoo_tests
   script:
     # print system info
     - which nvidia-smi

niftynet/application/base_application.py

@@ -18,16 +18,22 @@
 EVAL = "evaluation"
 
 
-class SingletonApplication(type):
-    _instances = None
 
+application_singleton_instance = None # global so it can be reset
+class SingletonApplication(type):
     def __call__(cls, *args, **kwargs):
-        if cls._instances is None:
-            cls._instances = \
+        global application_singleton_instance
+        if application_singleton_instance is None:
+            application_singleton_instance = \
                 super(SingletonApplication, cls).__call__(*args, **kwargs)
         # else:
         #     raise RuntimeError('application instance already started.')
-        return cls._instances
+        return application_singleton_instance
+
+    @classmethod
+    def clear(cls):
+        global application_singleton_instance
+        application_singleton_instance = None
 
 
 class BaseApplication(with_metaclass(SingletonApplication, object)):

niftynet/engine/application_driver.py

@@ -507,11 +507,11 @@ def _device_string(self, device_id=0, is_worker=True):
             # in training: use gpu only for workers whenever n_local_gpus
             device = 'gpu' if (is_worker and n_local_gpus > 0) else 'cpu'
             if device == 'gpu' and device_id >= n_local_gpus:
-                tf.logging.fatal(
+                tf.logging.warning(
                     'trying to use gpu id %s, but only has %s GPU(s), '
                     'please set num_gpus to %s at most',
                     device_id, n_local_gpus, n_local_gpus)
-                raise ValueError
+                #raise ValueError
             return '/{}:{}'.format(device, device_id)
         # in inference: use gpu for everything whenever n_local_gpus
         return '/gpu:0' if n_local_gpus > 0 else '/cpu:0'

niftynet/engine/windows_aggregator_base.py

@@ -24,6 +24,7 @@ def __init__(self, image_reader=None, output_path='.'):
         self._image_id = None
         self.prefix = ''
         self.output_path = os.path.abspath(output_path)
+        self.inferred_cleared = False
 
     @property
     def input_image(self):
@@ -133,3 +134,22 @@ def crop_batch(window, location, border):
                 ' spatial dims are: %s', window_shape, spatial_shape)
             raise NotImplementedError
         return window, location
+
+    def log_inferred(self, subject_name, filename):
+        """
+        This function writes out a csv of inferred files
+
+        :param subject_name: subject name corresponding to output
+        :param filename: filename of output
+        :return:
+        """
+        inferred_csv = os.path.join(self.output_path, 'inferred.csv')
+        if not self.inferred_cleared:
+            if os.path.exists(inferred_csv):
+                os.remove(inferred_csv)
+            self.inferred_cleared = True
+            if not os.path.exists(self.output_path):
+                os.makedirs(self.output_path)
+        with open(inferred_csv, 'a+') as csv_file:
+            filename = os.path.join(self.output_path, filename)
+            csv_file.write('{},{}\n'.format(subject_name, filename))

niftynet/engine/windows_aggregator_classifier.py

@@ -30,11 +30,7 @@ def __init__(self,
         self.name = name
         self.output_interp_order = 0
         self.prefix = prefix
-        self.output_path = os.path.abspath(output_path)
-        self.inferred_csv = os.path.join(self.output_path, 'inferred.csv')
         self.csv_path = os.path.join(self.output_path, self.prefix+'.csv')
-        if os.path.exists(self.inferred_csv):
-            os.remove(self.inferred_csv)
         if os.path.exists(self.csv_path):
             os.remove(self.csv_path)
 
@@ -72,7 +68,5 @@ def _save_current_image(self, image_out):
         with open(self.csv_path, 'a') as csv_file:
             data_str = ','.join([str(i) for i in image_out[0, 0, 0, 0, :]])
             csv_file.write(subject_name+','+data_str+'\n')
-        with open(self.inferred_csv, 'a') as csv_file:
-            filename = os.path.join(self.output_path, filename)
-            csv_file.write('{},{}\n'.format(subject_name, filename))
+        self.log_inferred(subject_name, filename)
         return

niftynet/engine/windows_aggregator_grid.py

@@ -33,13 +33,9 @@ def __init__(self,
             self, image_reader=image_reader, output_path=output_path)
         self.name = name
         self.image_out = None
-        self.output_path = os.path.abspath(output_path)
-        self.inferred_csv = os.path.join(self.output_path, 'inferred.csv')
         self.window_border = window_border
         self.output_interp_order = interp_order
         self.prefix = prefix
-        if os.path.exists(self.inferred_csv):
-            os.remove(self.inferred_csv)
 
     def decode_batch(self, window, location):
         n_samples = location.shape[0]
@@ -91,7 +87,5 @@ def _save_current_image(self):
                                 self.image_out,
                                 source_image_obj,
                                 self.output_interp_order)
-        with open(self.inferred_csv, 'a') as csv_file:
-            filename = os.path.join(self.output_path, filename)
-            csv_file.write('{},{}\n'.format(subject_name, filename))
+        self.log_inferred(subject_name, filename)
         return

niftynet/engine/windows_aggregator_resize.py

@@ -32,13 +32,9 @@ def __init__(self,
         ImageWindowsAggregator.__init__(
             self, image_reader=image_reader, output_path=output_path)
         self.name = name
-        self.output_path = os.path.abspath(output_path)
-        self.inferred_csv = os.path.join(self.output_path, 'inferred.csv')
         self.window_border = window_border
         self.output_interp_order = interp_order
         self.prefix = prefix
-        if os.path.exists(self.inferred_csv):
-            os.remove(self.inferred_csv)
 
     def decode_batch(self, window, location):
         """
@@ -103,7 +99,5 @@ def _save_current_image(self, image_out, resize_to):
                                 image_out,
                                 source_image_obj,
                                 self.output_interp_order)
-        with open(self.inferred_csv, 'a') as csv_file:
-            filename = os.path.join(self.output_path, filename)
-            csv_file.write('{},{}\n'.format(subject_name, filename))
+        self.log_inferred(subject_name, filename)
         return

niftynet/layer/resampler.py

@@ -466,10 +466,12 @@ def _binary_neighbour_ids(spatial_rank):
             for i in range(2 ** spatial_rank)]
 
 
-@tf.RegisterGradient('FloorMod')
-def _floormod_grad(op, grad):
-    return [None, None]
-
+try: # Some tf versions have this defined already
+    @tf.RegisterGradient('FloorMod')
+    def _floormod_grad(op, grad):
+        return [None, None]
+except:
+    pass
 
 SUPPORTED_INTERPOLATION = {'BSPLINE', 'LINEAR', 'NEAREST', 'IDW'}
 

niftynet/utilities/download.py

@@ -41,11 +41,13 @@ def download(example_ids,
         A list of identifiers for the samples to download
     :param download_if_already_existing:
         If true, data will always be downloaded
+    :param verbose:
+        If true, download info will be printed
     """
 
     global_config = NiftyNetGlobalConfig()
 
-    config_store = ConfigStore(global_config)
+    config_store = ConfigStore(global_config, verbose=verbose)
 
     # If a single id is specified, convert to a list
     example_ids = [example_ids] \
@@ -111,7 +113,7 @@ def download_file(url, download_path):
     move(downloaded_file, destination_path)
 
 
-def download_and_decompress(url, download_path):
+def download_and_decompress(url, download_path, verbose=True):
     """
     Download an archive from a resource URL and
     decompresses/unarchives to the given location
@@ -132,7 +134,10 @@ def download_and_decompress(url, download_path):
     downloaded_file = os.path.join(tempfile.gettempdir(), filename)
 
     # Download the file
-    urlretrieve(url, downloaded_file, reporthook=progress_bar_wrapper)
+    if verbose:
+        urlretrieve(url, downloaded_file, reporthook=progress_bar_wrapper)
+    else:
+        urlretrieve(url, downloaded_file)
 
     # Decompress and extract all files to the specified local path
     tar = tarfile.open(downloaded_file, "r")
@@ -149,13 +154,14 @@ class ConfigStore:
     remote repository with local caching
     """
 
-    def __init__(self, global_config):
+    def __init__(self, global_config, verbose=True):
         self._download_folder = global_config.get_niftynet_home_folder()
         self._config_folder = global_config.get_niftynet_config_folder()
         self._local = ConfigStoreCache(
             os.path.join(self._config_folder, '.downloads_local_config_cache'))
         self._remote = RemoteProxy(self._config_folder,
                                    global_config.get_download_server_url())
+        self._verbose = verbose
 
     def exists(self, example_id):
         """
@@ -258,8 +264,10 @@ def _download(self, remote_config_sections, example_id):
                                          'configuration file')
                     local_download_path = self._get_local_download_path(
                         config_params, example_id)
-                    download_and_decompress(url=config_params['url'],
-                                            download_path=local_download_path)
+                    download_and_decompress(
+                        url=config_params['url'],
+                        download_path=local_download_path,
+                        verbose=self._verbose)
                     print('{} -- {}: OK.'.format(example_id, section_name))
                     print("Downloaded data to " + local_download_path)
                 else:

tests/rand_flip_test.py

@@ -8,8 +8,6 @@
 
 from niftynet.layer.rand_flip import RandomFlipLayer
 
-os.environ["CUDA_VISIBLE_DEVICES"] = '-1'
-
 
 class RandFlipTest(tf.test.TestCase):
     def test_1d_flip(self):