Fix: multidimensional mutual information

Author: LauritzR

setup.py

@@ -5,7 +5,7 @@
 
 setuptools.setup(
     name="principal-feature-analysis",
-    version="1.0.8",
+    version="1.0.9",
     author="Tim Breitenbach & Lauritz Rasbach",
     author_email="[email protected], [email protected]",
     description="The first package for Principal Feature Analysis",

src/principal_feature_analysis/get_mutual_information.py

@@ -81,7 +81,7 @@ def make_summand_from_frequencies(x,y):
     for i in range(0,number_output_functions):
         list_of_features=list(range(number_output_functions,len(left_features)))
         list_of_features.insert(0,i)
-        id_features=np.array(list_variables)[left_features]
+        id_features=np.array(list_variables)[list_of_features]
 
         for j in list_of_features:
             freq_data_product = ((np.histogram2d(data[i,:], data[left_features[j], :], bins=(l[i], l[left_features[j]]))[0])) / n