Download data from OSF for testing

aroma/features.py

@@ -6,7 +6,7 @@
 from nilearn import image, masking
 from nilearn._utils import load_niimg
 
-from . import utils
+from aroma import utils
 
 LGR = logging.getLogger(__name__)
 
@@ -79,23 +79,14 @@ def feature_time_series(mel_mix, mc, metric_metadata=None):
 
     # Determine the derivatives of the RPs (add zeros at time-point zero)
     _, nparams = rp6.shape
-    rp6_der = np.vstack((
-        np.zeros(nparams),
-        np.diff(rp6, axis=0)
-    ))
+    rp6_der = np.vstack((np.zeros(nparams), np.diff(rp6, axis=0)))
 
     # Create an RP-model including the RPs and its derivatives
     rp12 = np.hstack((rp6, rp6_der))
 
     # add the fw and bw shifted versions
-    rp12_1fw = np.vstack((
-        np.zeros(2 * nparams),
-        rp12[:-1]
-    ))
-    rp12_1bw = np.vstack((
-        rp12[1:],
-        np.zeros(2 * nparams)
-    ))
+    rp12_1fw = np.vstack((np.zeros(2 * nparams), rp12[:-1]))
+    rp12_1bw = np.vstack((rp12[1:], np.zeros(2 * nparams)))
     rp_model = np.hstack((rp12, rp12_1fw, rp12_1bw))
 
     # Determine the maximum correlation between RPs and IC time-series
@@ -109,9 +100,7 @@ def feature_time_series(mel_mix, mc, metric_metadata=None):
     for i in range(nsplits):
         # Select a random subset of 90% of the dataset rows
         # (*without* replacement)
-        chosen_rows = np.random.choice(a=range(nmixrows),
-                                       size=nrows_to_choose,
-                                       replace=False)
+        chosen_rows = np.random.choice(a=range(nmixrows), size=nrows_to_choose, replace=False)
 
         # Combined correlations between RP and IC time-series, squared and
         # non squared
@@ -293,8 +282,9 @@ def feature_spatial(mel_IC, metric_metadata=None):
         tot_sum = np.sum(temp_IC_data)
 
         if tot_sum == 0:
-            LGR.info("\t- The spatial map of component {} is empty. "
-                     "Please check!".format(i + 1))
+            LGR.info(
+                "\t- The spatial map of component {} is empty. " "Please check!".format(i + 1)
+            )
 
         # Get sum of Z-values of the voxels located within the CSF
         # (calculate via the mean and number of non-zero voxels)

aroma/tests/conftest.py

@@ -1,7 +1,8 @@
-import os.path as op
+import os
+import ssl
+from urllib.request import urlretrieve
 
 import aroma
-import numpy as np
 import pytest
 from nilearn.datasets import fetch_development_fmri
 
@@ -20,83 +21,126 @@ def skip_integration(request):
     return request.config.getoption("--skipintegration")
 
 
+def fetch_file(osf_id, path, filename):
+    """
+    Fetches file located on OSF and downloads to `path`/`filename`1
+
+    Parameters
+    ----------
+    osf_id : str
+        Unique OSF ID for file to be downloaded. Will be inserted into relevant
+        location in URL: https://osf.io/{osf_id}/download
+    path : str
+        Path to which `filename` should be downloaded. Ideally a temporary
+        directory
+    filename : str
+        Name of file to be downloaded (does not necessarily have to match name
+        of file on OSF)
+
+    Returns
+    -------
+    full_path : str
+        Full path to downloaded `filename`
+    """
+    # This restores the same behavior as before.
+    # this three lines make tests dowloads work in windows
+    if os.name == "nt":
+        orig_sslsocket_init = ssl.SSLSocket.__init__
+        ssl.SSLSocket.__init__ = (
+            lambda *args, cert_reqs=ssl.CERT_NONE, **kwargs: orig_sslsocket_init(
+                *args, cert_reqs=ssl.CERT_NONE, **kwargs
+            )
+        )
+        ssl._create_default_https_context = ssl._create_unverified_context
+    url = "https://osf.io/{}/download".format(osf_id)
+    full_path = os.path.join(path, filename)
+    if not os.path.isfile(full_path):
+        urlretrieve(url, full_path)
+    return full_path
+
+
 @pytest.fixture(scope="session")
 def testpath(tmp_path_factory):
     """Test path that will be used to download all files."""
-    return tmp_path_factory.mktemp(basename='data', numbered=False)
+    return tmp_path_factory.mktemp(basename="data", numbered=False)
 
 
 @pytest.fixture
-def base_test_path():
-    return op.join(op.dirname(aroma.__file__), 'tests', 'data')
+def nilearn_data(testpath):
+    return fetch_development_fmri(n_subjects=1, age_group="adult", data_dir=str(testpath))
+
+
+# Feature outputs generated with the following command (adding breakpoints to save results)
+# python2 ICA_AROMA.py -o out -i
+# sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz
+# -mc mc.tsv -tr 2 -np
 
 
 @pytest.fixture
-def featurespath(base_test_path):
-    return op.join(base_test_path, 'features_test')
+def mel_FT_mix(testpath):
+    return fetch_file("ezhfg", testpath, "melodic_FTmix")
 
 
 @pytest.fixture
-def nilearn_data(testpath):
-    return fetch_development_fmri(n_subjects=1, age_group="adult",
-                                  data_dir=str(testpath))
+def mel_mix(testpath):
+    return fetch_file("69j2h", testpath, "melodic_mix")
 
-# Feature outputs generated with the following command (adding breakpoints to save results)
-# python2 ICA_AROMA.py -o out -i
-# sub-pixar123_task-pixar_space-MNI152NLin2009cAsym_desc-preproc_bold.nii.gz
-# -mc mc.tsv -tr 2 -np
+
+@pytest.fixture
+def mc(testpath):
+    return fetch_file("af275", testpath, "mc_fmriprep.tsv")
 
 
 @pytest.fixture
-def mel_FT_mix(featurespath):
-    return np.loadtxt(op.join(featurespath, 'melodic_FTmix'))
+def mel_IC(testpath):
+    return fetch_file("bw3mq", testpath, "melodic_IC_thr_MNI2mm.nii.gz")
 
 
 @pytest.fixture
-def mel_mix(featurespath):
-    return np.loadtxt(op.join(featurespath, 'melodic_mix'))
+def csfFract(testpath):
+    return fetch_file("d298c", testpath, "csfFract.npy")
 
 
 @pytest.fixture
-def mc(featurespath):
-    return op.join(featurespath, 'mc.txt')
+def edgeFract(testpath):
+    return fetch_file("vxdu6", testpath, "edgeFract.npy")
 
 
 @pytest.fixture
-def mel_IC(featurespath):
-    return op.join(featurespath, 'melodic_IC_thr_MNI2mm.nii.gz')
+def max_correls(testpath):
+    return fetch_file("hpd2m", testpath, "max_correls.npy")
 
 
 @pytest.fixture
-def csfFract(featurespath):
-    return op.join(featurespath, 'csfFract.npy')
+def HFC(testpath):
+    return fetch_file("rymwg", testpath, "HFC.npy")
 
 
 @pytest.fixture
-def edgeFract(featurespath):
-    return op.join(featurespath, 'edgeFract.npy')
+def classification_overview(testpath):
+    return fetch_file("2hv3a", testpath, "classification_overview.txt")
 
 
 @pytest.fixture
-def max_correls(featurespath):
-    return op.join(featurespath, 'max_correls.npy')
+def classified_motion_ICs(testpath):
+    return fetch_file("jnhgc", testpath, "AROMAnoiseICs.csv")
 
 
 @pytest.fixture
-def HFC(featurespath):
-    return op.join(featurespath, 'HFC.npy')
+def feature_scores(testpath):
+    return fetch_file("cxwfk", testpath, "feature_scores.txt")
 
 
 @pytest.fixture
-def motion_parameters(featurespath):
+def motion_parameters(testpath):
     """Motion parameter outputs in different formats.
 
     All outputs manually converted from FSL version.
     """
     files = {
-        "FSL": op.join(featurespath, "mc_fsl.txt"),
-        "AfNI": op.join(featurespath, "mc_afni.1D"),
-        "SPM": op.join(featurespath, "rp_mc_spm.txt"),
-        "fMRIPrep": op.join(featurespath, "mc_fmriprep.tsv"),
+        "FSL": fetch_file("ahtrv", testpath, "mc_fsl.txt"),
+        "AfNI": fetch_file("p6ybt", testpath, "mc_afni.1D"),
+        "SPM": fetch_file("ct6q4", testpath, "rp_mc_spm.txt"),
+        "fMRIPrep": fetch_file("af275", testpath, "mc_fmriprep.tsv"),
     }
     return files