fix testing paths

BrainNetworksInPython/scripts/make_corr_matrices.py

@@ -2,14 +2,10 @@
 """
 Tools to create a correlation matrix from regional measures
 """
-# A Global import to make code python 2 and 3 compatible
-from __future__ import print_function
-
-# Other essential package imports
+# Essential package imports
 import os
 import numpy as np
 import pandas as pd
-import stats_functions
 
 def get_non_numeric_cols(df):
     numeric = np.fromiter((np.issubdtype(y, np.number) for y in df.dtypes),bool)
@@ -26,6 +22,8 @@ def create_residuals_df(df, names, covars_list):
       you choose to correct for before correlating the regions.
     df should be numeric for the columns in names and covars_list
     '''
+    import BrainNetworksInPython.scripts.stats_functions as sf
+
     # Raise TypeError if any of the relevant columns are nonnumeric
     non_numeric_cols = get_non_numeric_cols(df[names+covars_list])
     if non_numeric_cols:
@@ -46,7 +44,7 @@ def create_residuals_df(df, names, covars_list):
 
     # Calculate the residuals
     for name in names:
-        df_res.loc[:, name] = stats_functions.residuals(x.T, df.loc[:, name])
+        df_res.loc[:, name] = sf.residuals(x.T, df.loc[:, name])
 
     # Return the residuals data frame
     return df_res

BrainNetworksInPython/scripts/stat_functions.py

@@ -0,0 +1,69 @@
+#!/usr/bin/env python
+
+# Essential package imports
+import numpy as np
+
+def residuals(x, y):
+    '''
+    Return residuals of least squares solution to y = AB where A =[[x 1]].
+    Uses numpy.linalg.lstsq to find B
+    '''
+
+    if len(x.shape) == 1:
+        x = x[np.newaxis, :]
+    A = np.vstack([x, np.ones(x.shape[-1])]).T
+    # get the least squares solution to AB = y
+    B = np.linalg.lstsq(A, y)[0]
+    # calculate and return the residuals
+    m, c = B[:-1], B[-1]
+    pre = np.sum(m * x.T, axis=1) + c
+    res = y - pre
+    return res
+
+def partial_r(x, y, covars):
+    from scipy.stats import pearsonr
+
+    res_i = residuals(covars, x)
+    res_j = residuals(covars, y)
+    part_r = pearsonr(res_i, res_j)[0]
+    return part_r
+
+def variance_partition(x1, x2, y):
+    '''
+    Describe the independent and shared explanatory
+    variance of two (possibly correlated) variables on 
+    the dependent variable (y)
+    '''
+    from statsmodels.formula.api import ols
+    from scipy.stats import pearsonr
+    import pandas as pd
+
+    # Set up the data frame
+    df = pd.DataFrame( { 'Y' : y ,
+                         'X1' : x1,
+                         'X2' : x2 } )
+
+    # Get the overall r squared value for the 
+    # multiple regression
+    Rsq = ols('Y ~ X1 + X2', data=df).fit().rsquared
+
+    # Next calculate the residuals of X1 after correlating
+    # with X2 (so units will be in those of X1) and vice versa
+    df['res_X1givenX2'] = residuals(df['X2'], df['X1'])
+    df['res_X2givenX1'] = residuals(df['X1'], df['X2'])
+
+    # Now calculate the pearson regressions for
+    # the residuals against the dependent variable to give
+    # the fraction of variance that each explains independently
+    # (a and c), along with the fraction of variance 
+    # that is shared across both explanatory variables (b).
+    # d is the fraction of variance that is not explained
+    # by the model.
+    a = (pearsonr(df['res_X1givenX2'], df['Y'])[0])**2
+    c = (pearsonr(df['res_X2givenX1'], df['Y'])[0])**2    
+    b = Rsq - a - c
+    d = 1.0 - Rsq
+
+    # Return these four fractions
+    return a, b, c, d
+

BrainNetworksInPython/scripts/stats_functions.py

@@ -1,11 +1,13 @@
 #!/usr/bin/env python
 
+# Essential package imports
+import numpy as np
+
 def residuals(x, y):
     '''
     Return residuals of least squares solution to y = AB where A =[[x 1]].
     Uses numpy.linalg.lstsq to find B
     '''
-    import numpy as np
 
     if len(x.shape) == 1:
         x = x[np.newaxis, :]
@@ -19,8 +21,6 @@ def residuals(x, y):
     return res
 
 def partial_r(x, y, covars):
-
-    import numpy as np
     from scipy.stats import pearsonr
 
     res_i = residuals(covars, x)
@@ -35,7 +35,6 @@ def variance_partition(x1, x2, y):
     the dependent variable (y)
     '''
     from statsmodels.formula.api import ols
-    import numpy as np
     from scipy.stats import pearsonr
     import pandas as pd
 
@@ -67,4 +66,4 @@ def variance_partition(x1, x2, y):
 
     # Return these four fractions
     return a, b, c, d
-
+

BrainNetworksInPython/tests/make_corr_matrices_test.py → tests/make_corr_matrices_test.py

@@ -1,11 +1,7 @@
 import os
 import sys
-sys.path.append(os.path.abspath(os.path.join('wrappers')))
-sys.path.append(os.path.abspath(os.path.join('example_data')))
-sys.path.append(os.path.abspath(os.path.join('scripts')))
-
-import BrainNetworksInPython.make_corr_matrices as mcm
-import BrainNetworksInPython.stats_functions as sf
+from BrainNetworksInPython.scripts import stats_functions as sf
+from BrainNetworksInPython.scripts import make_corr_matrices as mcm
 import pytest
 import pandas as pd
 import numpy as np

tests/make_graphs_test.py

@@ -0,0 +1,9 @@
+import os
+import sys
+
+import pytest
+import pandas as pd
+import numpy as np
+
+import BrainNetworksInPython.scripts.make_graphs as mkg
+

BrainNetworksInPython/tests/regression_test.py → tests/regression_test.py

@@ -1,5 +1,6 @@
 import unittest
 import os
+import sys
 
 class FixturesTest(unittest.TestCase):
 
@@ -8,7 +9,7 @@ class FixturesTest(unittest.TestCase):
     def setUpClass(cls):
         print('\nin set up - this takes about 80 secs')
 
-        from BrainNetworksInPython.tests.write_fixtures import generate_fixture_hashes, unpickle_hash
+        from tests.write_fixtures import generate_fixture_hashes, unpickle_hash
         cls.hash_dict_new = generate_fixture_hashes()
         cls.hash_dict_original = unpickle_hash()
         # define dictionary keys for individual files for checking
@@ -17,7 +18,6 @@ def setUpClass(cls):
         cls.gm = folder + '/network-analysis/GlobalMeasures_corrmat_file_COST010.csv'
         cls.lm = folder + '/network-analysis/NodalMeasures_corrmat_file_COST010.csv'
         cls.rich = folder + '/network-analysis/RICH_CLUB_corrmat_file_COST010.csv'
-        import 
 
     #--------------------------- Tests --------------------------------
     # Each of these tests checks that ourly newly generated version of

BrainNetworksInPython/tests/write_fixtures.py → tests/write_fixtures.py

@@ -1,12 +1,10 @@
 #--------------------------- Write fixtures ---------------------------
 # To regression test our wrappers we need examples. This script 
-# generates files. We save these files once, and regression-tests.py
+# generates files. We save these files once, and regression_test.py
 # re-generates these files to tests them for identicality with the
 # presaved examples (fixtures). If they are found not to be identical 
 # it throws up an error. 
 #
-# (*someday* I would like to replace saving files with saving hashes)
-#
 # The point of this is to check that throughout the changes we make to 
 # BrainNetworksInPython the functionality of this script stays the same
 #
@@ -23,7 +21,7 @@ def recreate_correlation_matrix_fixture(folder):
     ##### the Whitaker_Vertes dataset                             ##### 
     import BrainNetworksInPython.datasets.NSPN_WhitakerVertes_PNAS2016.data as data
     centroids, regionalmeasures, names, covars, names_308_style = data._get_data()
-    from BrainNetworksInPython.corrmat_from_regionalmeasures import corrmat_from_regionalmeasures
+    from BrainNetworksInPython.wrappers.corrmat_from_regionalmeasures import corrmat_from_regionalmeasures
     corrmat_path = os.getcwd()+folder+'/corrmat_file.txt'
     corrmat_from_regionalmeasures(
         regionalmeasures,
@@ -42,7 +40,7 @@ def recreate_network_analysis_fixture(folder, corrmat_path):
     # calculate global measures
     import random
     random.seed(2984)
-    from BrainNetworksInPython.network_analysis_from_corrmat import network_analysis_from_corrmat
+    from BrainNetworksInPython.wrappers.network_analysis_from_corrmat import network_analysis_from_corrmat
     network_analysis_from_corrmat(corrmat_path,
                               names,
                               centroids,

tutorials/jupyter_demo.ipynb

@@ -42,7 +42,9 @@
   {
    "cell_type": "code",
    "execution_count": 1,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "import sys\n",
@@ -205,7 +207,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "# We will be using the files we just used network_analysis_from _corrmat to create\n",
@@ -238,7 +242,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.6.4"
+   "version": "3.6.3"
   }
  },
  "nbformat": 4,