update deprecated binom_test (#426)

* updated deprecated function name

* fix roc print output

* fix roc print output

* fix path check test on windows

* Remove sphinx napoleon as it's included with sphinx now

* Fix random seeds for reproducible testing

---------

Co-authored-by: ejolly <[email protected]>

docs/conf.py

@@ -44,7 +44,6 @@
     "sphinx.ext.autosummary",
     "sphinx.ext.viewcode",
     "sphinx.ext.mathjax",
-    "sphinxcontrib.napoleon",
     "sphinx_gallery.gen_gallery",
 ]
 

nltools/analysis.py

@@ -12,13 +12,12 @@
 import pandas as pd
 import numpy as np
 from nltools.plotting import roc_plot
-from scipy.stats import norm, binom_test
+from scipy.stats import norm, binomtest
 from sklearn.metrics import auc
 from copy import deepcopy
 
 
 class Roc(object):
-
     """Roc Class
 
     The Roc class is based on Tor Wager's Matlab roc_plot.m function and
@@ -233,7 +232,7 @@ def calculate(
 
         # Calculate p-Value using binomial test (can add hierarchical version of binomial test)
         self.n = len(self.misclass)
-        self.accuracy_p = binom_test(int(np.sum(~self.misclass)), self.n, p=0.5)
+        self.accuracy_p = binomtest(int(np.sum(~self.misclass)), self.n, p=0.5)
         self.accuracy_se = np.sqrt(
             np.mean(~self.misclass) * (np.mean(~self.misclass)) / self.n
         )
@@ -318,7 +317,10 @@ def summary(self):
         print("------------------------")
         print("{:20s}".format("Accuracy:") + "{:.2f}".format(self.accuracy))
         print("{:20s}".format("Accuracy SE:") + "{:.2f}".format(self.accuracy_se))
-        print("{:20s}".format("Accuracy p-value:") + "{:.2f}".format(self.accuracy_p))
+        print(
+            "{:20s}".format("Accuracy p-value:")
+            + "{:.2f}".format(self.accuracy_p.pvalue)
+        )
         print("{:20s}".format("Sensitivity:") + "{:.2f}".format(self.sensitivity))
         print("{:20s}".format("Specificity:") + "{:.2f}".format(self.specificity))
         print("{:20s}".format("AUC:") + "{:.2f}".format(self.auc))

nltools/tests/conftest.py

@@ -10,6 +10,7 @@
 
 @pytest.fixture(scope="module", params=["2mm"])
 def sim_brain_data():
+    np.random.seed(0)
     # MNI_Template["resolution"] = request.params
     sim = Simulator()
     r = 10
@@ -25,6 +26,7 @@ def sim_brain_data():
 
 @pytest.fixture(scope="module")
 def sim_design_matrix():
+    np.random.seed(0)
     # Design matrices are specified in terms of sampling frequency
     TR = 2.0
     sampling_freq = 1.0 / TR
@@ -37,6 +39,7 @@ def sim_design_matrix():
 
 @pytest.fixture(scope="module")
 def sim_adjacency_single():
+    np.random.seed(0)
     sim = np.random.multivariate_normal(
         [0, 0, 0, 0],
         [
@@ -54,6 +57,7 @@ def sim_adjacency_single():
 
 @pytest.fixture(scope="module")
 def sim_adjacency_multiple():
+    np.random.seed(0)
     n = 10
     sim = np.random.multivariate_normal(
         [0, 0, 0, 0],

nltools/tests/test_brain_data.py

@@ -75,7 +75,9 @@ def test_load(tmpdir):
         bb = Brain_Data(
             os.path.join(tmpdir.join("test_write.h5")), mask=MNI_Template["mask"]
         )
-        assert bb.mask.get_filename() == MNI_Template["mask"]
+        assert os.path.abspath(bb.mask.get_filename()) == os.path.abspath(
+            MNI_Template["mask"]
+        )
 
 
 def test_shape(sim_brain_data):

requirements-dev.txt

@@ -6,5 +6,3 @@ coveralls
 sphinx
 sphinx_gallery
 sphinx_bootstrap_theme
-sphinxcontrib-napoleon
-

requirements.txt

@@ -5,7 +5,7 @@ pandas>=1.1.0
 numpy<1.24
 seaborn>=0.7.0
 matplotlib>=2.2.0
-scipy
+scipy>=1.7.0
 pynv
 joblib>=0.15
 h5py