Merge pull request #52 from AlonDaks/alon-dev

Packaged our code as a python module called stat159lambda

.gitignore

@@ -1,5 +1,7 @@
 .DS_Store
 
+# Ignore Data files
+.nii
 
 # Byte-compiled / optimized / DLL files
 __pycache__/

Makefile

@@ -6,10 +6,10 @@ clean:
 	find . -name "*.so" -o -name "*.pyc" -o -name "*.pyx.md5" | xargs rm -f
 
 coverage:
-	nosetests code/utils data --with-coverage --cover-package=data  --cover-package=utils
+	nosetests code/stat159lambda/utils data --with-coverage --cover-package=data  --cover-package=utils
 
 test:
-	nosetests code/utils data
+	nosetests code/stat159lambda/utils data
 
 verbose:
-	nosetests -v code/utils data
+	nosetests -v code/stat159lambda/utils data

code/stat159lambda/config.py

@@ -0,0 +1,12 @@
+import os
+import sys
+
+REPO_HOME_PATH = os.path.realpath(__file__).replace('/code/stat159lambda/config.py', '')
+
+USE_CACHED_DATA = os.environ.get('STAT159_CACHED_DATA', 'True') == 'True'
+
+NUM_PROCESSES = int(os.environ.get('STAT159_NUM_PROCESSES', 5))
+
+NUM_VOXELS = 1108800
+
+NUM_TOTAL_VOLUMES = 3543

code/reproduction/inter_run_diagnostics.py → code/stat159lambda/reproduction/inter_run_diagnostics.py

@@ -4,9 +4,7 @@
 import sys
 import matplotlib.pyplot as plt
 import gc
-
-REPO_HOME_RELATIVE_PATH = '../../'
-sys.path.append(REPO_HOME_RELATIVE_PATH)
+from stat159lambda.config import REPO_HOME_PATH
 
 
 def calc_vol_rms_diff(data_file_path):
@@ -20,14 +18,14 @@ def calc_vol_rms_diff(data_file_path):
 
 def save_plot(vol_rms_diff, subj_num):
     plt.plot(vol_rms_diff)
-    plt.savefig('{0}figures/subj{1}_vol_rms_diff.png'.format(
-        REPO_HOME_RELATIVE_PATH, subj_num))
+    plt.savefig('{0}/figures/subj{1}_vol_rms_diff.png'.format(
+        REPO_HOME_PATH, subj_num))
 
 
 if __name__ == '__main__':
     subj_num = sys.argv[1]
-    data_file_path = '{0}data/processed/sub{1}_rcds_2d.npy'.format(
-        REPO_HOME_RELATIVE_PATH, subj_num)
+    data_file_path = '{0}/data/processed/sub{1}_rcds_2d.npy'.format(
+        REPO_HOME_PATH, subj_num)
     vol_rms_diff = calc_vol_rms_diff(data_file_path)
     save_plot(vol_rms_diff, subj_num)
     del vol_rms_diff

code/reproduction/preprocess.py → code/stat159lambda/reproduction/preprocess.py

@@ -7,26 +7,23 @@
 from glob import glob
 from os.path import exists
 from multiprocessing import Pool
+from stat159lambda.config import REPO_HOME_PATH
 
-REPO_HOME_RELATIVE_PATH = '../../'
-sys.path.append(REPO_HOME_RELATIVE_PATH)
-import config as cf
-
-DATA_PATHS = json.load(open(REPO_HOME_RELATIVE_PATH + 'data/data_path.json'))
+DATA_PATHS = json.load(open('{0}/data/data_path.json'.format(REPO_HOME_PATH)))
 
 INCORRECT_NUM_ARGS_MESSAGE = 'Invalid number of arguments: specify alignment type'
 ILLEGAL_ARG_MESSAGE = 'preprocess.py must be provided with alignment argument'
 
 
 def concatenate_runs(alignment):
     for i, subject in enumerate(DATA_PATHS['subjects']):
-        nii_file_name = '{0}data/processed/sub{1}_{2}'.format(
-            REPO_HOME_RELATIVE_PATH, i + 1, alignment)
+        nii_file_name = '{0}/data/processed/sub{1}_{2}'.format(
+            REPO_HOME_PATH, i + 1, alignment)
         if not exists(nii_file_name + '.nii') or not cf.USE_CACHED_DATA:
             run_data = []
             for j, run in enumerate(subject['runs']):
                 task_path = run[alignment]['path']
-                img = nib.load(REPO_HOME_RELATIVE_PATH + task_path)
+                img = nib.load(REPO_HOME_PATH + '/' + task_path)
                 data = img.get_data()
                 if j == 0:
                     run_data.append(data[..., :-4])
@@ -44,8 +41,8 @@ def concatenate_runs(alignment):
 
 
 def reshape_data_to_2d(alignment):
-    files_to_reshape = np.sort(glob('{0}data/processed/sub*_{1}.nii'.format(
-        REPO_HOME_RELATIVE_PATH, alignment)))
+    files_to_reshape = np.sort(glob('{0}/data/processed/sub*_{1}.nii'.format(
+        REPO_HOME_PATH, alignment)))
     for f in files_to_reshape:
         file_name_2d = f.replace('.nii', '_2d')
         if not exists(file_name_2d + '.npy') or not cf.USE_CACHED_DATA:

code/reproduction/similarity.py → code/stat159lambda/reproduction/similarity.py

@@ -9,10 +9,7 @@
 from os import path
 import sharedmem as sm
 import gc
-
-REPO_HOME_RELATIVE_PATH = '../../'
-sys.path.append(REPO_HOME_RELATIVE_PATH)
-import config as cf
+from stat159lambda.config import REPO_HOME_PATH
 
 INCORRECT_NUM_ARGS_MESSAGE = 'invalid number of arguments: specify alignment type, subject1 and subject2'
 ILLEGAL_ALIGNMENT_ARG_MESSAGE = 'alignment argument must be either <rcds>, <linear>, <non_linear>'
@@ -54,10 +51,10 @@ def check_command_line_arguments(arguments):
         subject2 = int(arguments[3])
     except:
         raise ValueError(ILLEGAL_SUBJECTS_ARG_MESSAGE)
-    subject1_file_path = '{0}data/processed/sub{1}_{2}_2d.npy'.format(
-        REPO_HOME_RELATIVE_PATH, subject1, alignment)
-    subject2_file_path = '{0}data/processed/sub{1}_{2}_2d.npy'.format(
-        REPO_HOME_RELATIVE_PATH, subject2, alignment)
+    subject1_file_path = '{0}/data/processed/sub{1}_{2}_2d.npy'.format(
+        REPO_HOME_PATH, subject1, alignment)
+    subject2_file_path = '{0}/data/processed/sub{1}_{2}_2d.npy'.format(
+        REPO_HOME_PATH, subject2, alignment)
     if not path.exists(subject1_file_path):
         raise ValueError(
             'Filing missing: {0}, run preprocess.py to generate necessary file'.format(
@@ -72,8 +69,8 @@ def check_command_line_arguments(arguments):
 if __name__ == '__main__':
     subject1_file_path, subject2_file_path = check_command_line_arguments(
         sys.argv)
-    correlation_file_name = '{0}data/processed/r_sub{1}_sub{2}_{3}'.format(
-        REPO_HOME_RELATIVE_PATH, sys.argv[2], sys.argv[3], sys.argv[1])
+    correlation_file_name = '{0}/data/processed/r_sub{1}_sub{2}_{3}'.format(
+        REPO_HOME_PATH, sys.argv[2], sys.argv[3], sys.argv[1])
     if not path.exists(correlation_file_name +
                        '.npy') or not cf.USE_CACHED_DATA:
         shared_subject1 = sm.copy(np.load(subject1_file_path))

code/utils/linear_modeling.py → code/stat159lambda/utils/linear_modeling.py

@@ -5,7 +5,7 @@
 import numpy.linalg as npl
 from scipy.stats import t as t_dist
 from scipy.ndimage import gaussian_filter
-from . import scene_slicer as ssm
+import stat159lambda.utils.scene_slicer as ssm
 import nibabel as nib
 
 

code/utils/tests/test_linear_modeling.py → code/stat159lambda/utils/tests/test_linear_modeling.py

@@ -12,7 +12,7 @@
 import numpy.linalg as npl
 from scipy.stats import t as t_dist
 
-from .. import linear_modeling
+from stat159lambda.utils import linear_modeling
 from numpy.testing import assert_equal, assert_almost_equal
 
 # Make an X, y

code/utils/tests/test_parse_demographics.py → code/stat159lambda/utils/tests/test_parse_demographics.py

@@ -1,5 +1,5 @@
 from __future__ import absolute_import
-from .. import parse_demographics
+from stat159lambda.utils import parse_demographics
 
 import os
 import csv

code/utils/tests/test_pearson_1d.py → code/stat159lambda/utils/tests/test_pearson_1d.py

@@ -28,7 +28,7 @@
 
 import numpy as np
 
-from .. import pearson
+from stat159lambda.utils import pearson
 
 from numpy.testing import assert_almost_equal
 

code/utils/tests/test_pearson_2d.py → code/stat159lambda/utils/tests/test_pearson_2d.py

@@ -10,7 +10,7 @@
 
 import numpy as np
 
-from .. import pearson
+from stat159lambda.utils import pearson
 
 from numpy.testing import assert_almost_equal
 

code/utils/tests/test_plot.py → code/stat159lambda/utils/tests/test_plot.py

@@ -1,5 +1,5 @@
 from __future__ import absolute_import
-from .. import plot
+from stat159lambda.utils import plot
 
 import numpy as np
 import matplotlib

code/stat159lambda/utils/tests/test_rf.py

@@ -0,0 +1,28 @@
+from stat159lambda.utils import rf
+import numpy as np
+from sklearn.ensemble import RandomForestClassifier
+try:
+    from mock import patch
+except:
+    from unittest.mock import patch
+
+
+def test_prepare():
+    X = np.array([[10, 8, 8], [1, 11, 8], [4, 1, 1], [5, 0, 3], [7, 9, 1],
+                  [1, 5, 8], [6, 3, 0], [7, 10, 3], [5, 4, 3], [7, 0, 8]])
+    y = np.array([1, 0, 1, 1, 0, 2, 1, 1, 0, 0])
+    return X, y
+
+
+def test_cv_rf_accuracy():
+    X, y = test_prepare()
+    assert rf.cv_rf_accuracy(X, y, 2, 2, 2, num_folds=5) > 0
+
+
+# @patch.object(RandomForestClassifier, '__init__')
+# def test_rf_accuracy(mock_rf_init):
+#     X, y = test_prepare()
+#     rf.rf_accuracy(X, y, 2, 2, 2, num_folds=5)
+#     assert_array_equal(mock_rf_init.call_args[0][0],
+#                        np.array([[10, 8, 8], [1, 11, 8], [4, 1, 1], [5, 0, 3], [7, 9, 1],
+#                   [1, 5, 8], [6, 3, 0], [7, 10, 3], [5, 4, 3], [7, 0, 8]]))

code/utils/tests/test_scene_slicer.py → code/stat159lambda/utils/tests/test_scene_slicer.py

@@ -1,5 +1,5 @@
 from __future__ import absolute_import, division, print_function
-from .. import scene_slicer
+from stat159lambda.utils import scene_slicer
 import os
 import nibabel as nib
 import csv

code/utils/tests/test_variance.py → code/stat159lambda/utils/tests/test_variance.py

@@ -1,5 +1,5 @@
 from __future__ import absolute_import
-from .. import variance
+from stat159lambda.utils import variance
 
 import numpy as np
 import nibabel as nib

tools/travis_tools.sh

@@ -1,5 +1,6 @@
 # Tools for working with travis-ci
 export WHEELHOUSE="http://travis-wheels.scikit-image.org/"
+export PYTHONPATH=$(pwd)/code
 
 retry () {
     # https://gist.github.com/fungusakafungus/1026804