__init__.py fixes dependencies

EDApy/__init__.py

@@ -11,5 +11,3 @@
 import warnings
 
 warnings.simplefilter(action='ignore', category=FutureWarning)
-from sklearn.cluster import KMeans
-

EDApy/optimization/multivariate/EDA_multivariate.py

@@ -5,9 +5,9 @@
 import rpy2.robjects.packages as rp
 import numpy as np
 
-from EDApy.optimization.multivariate.__BayesianNetwork import learn_structure, calculate_fit
-from EDApy.optimization.multivariate.__matrix import nearestPD, normalizacion, is_invertible
-from EDApy.optimization.multivariate.__clustering import clustering
+from .__BayesianNetwork import learn_structure, calculate_fit
+from .__clustering import clustering
+from .__matrix import nearestPD, normalizacion, is_invertible
 
 utils = rp.importr('utils')
 utils.chooseCRANmirror(ind=1)

EDApy/optimization/multivariate/__clustering.py

@@ -3,7 +3,6 @@
 
 import numpy as np
 import pandas as pd
-from sklearn.cluster import KMeans
 
 
 def clustering(n_clusters, dataset, evidences, cluster_vars):
@@ -15,6 +14,7 @@ def clustering(n_clusters, dataset, evidences, cluster_vars):
 
     values = list(dataset.index)
 
+    from sklearn.cluster import KMeans
     k_means = KMeans(n_clusters=n_clusters, random_state=0).fit(np.array(data).T)
     labels = list(k_means.labels_)
 

EDApy/optimization/multivariate/__init__.py

@@ -3,7 +3,44 @@
 
 # __init__.py
 
-from .EDA_multivariate import EDAgbn as EDA_multivariate
-from .__BayesianNetwork import print_structure as print_structure
+# __all__ = ['EDA_multivariate', '__BayesianNetwork']
+
+import subprocess
+import sys
+
+import warnings
+warnings.filterwarnings("ignore")
+
+
+def install(package):
+    subprocess.check_call([sys.executable, "-m", "pip", "install", package])
+
 
+def check_package(installed_pack, package):
+    if package in installed_pack:
+        return True
+    else:
+        return False
 
+if sys.version_info[0] < 3:
+    raise Exception("Python version should be greater than 3")
+
+requirements = ['pandas', 'numpy', 'rpy2', 'pyvis', 'sklearn']
+reqs = subprocess.check_output([sys.executable, '-m', 'pip', 'freeze'])
+installed_packages = [r.decode().split('==')[0] for r in reqs.split()]
+
+for package in requirements:
+    if not check_package(installed_packages, package):
+        if package == 'sklearn':
+            package = 'scikit-learn'
+
+        print('package ', package, ' is not installed. Would you like to install it? y/n')
+        response = input()
+        if response == 'y' or response == 'yes':
+            install(package)
+        elif response == 'n' or response == 'no':
+            print('dependencies insatisfied')
+
+
+from .EDA_multivariate import EDAgbn as EDA_multivariate
+from .__BayesianNetwork import print_structure as print_structure

EDApy/optimization/univariate/__init__.py

@@ -3,6 +3,39 @@
 
 # __init__.py
 
+
+import subprocess
+import sys
+
+
+def install(package):
+    subprocess.check_call([sys.executable, "-m", "pip", "install", package])
+
+
+def check_package(installed_pack, package):
+    if package in installed_pack:
+        return True
+    else:
+        return False
+
+
+if sys.version_info[0] < 3:
+    raise Exception("Python version should be greater than 3")
+
+requirements = ['pandas', 'numpy', 'scipy']
+reqs = subprocess.check_output([sys.executable, '-m', 'pip', 'freeze'])
+installed_packages = [r.decode().split('==')[0] for r in reqs.split()]
+
+for package in requirements:
+    if not check_package(installed_packages, package):
+        print('package ', package, ' is not installed. Would you like to install it? y/n')
+        response = input()
+        if response == 'y' or response == 'yes':
+            install(package)
+        elif response == 'n' or response == 'no':
+            print('dependencies not satisfied')
+
+
 from .discrete import UMDAd as EDA_discrete
 from .continuous import UMDAc as EDA_continuous
 

EDApy/optimization/univariate/continuous.py

@@ -1,6 +1,5 @@
 import pandas as pd
 import numpy as np
-from scipy.stats import norm
 
 '''
 In this version of UMDA, instead of a vector of probabilities, a vector of univariate normal distributions is found
@@ -161,6 +160,7 @@ def update_vector(self):
             array = self.generation[var].values
 
             # calculate mu and std from data
+            from scipy.stats import norm
             mu, std = norm.fit(array)
 
             # std should never be 0
@@ -220,4 +220,3 @@ def run(self, output=True):
                 print('IT ', i, 'best cost ', best_mae_local)
 
         return self.best_mae_global, self.best_ind_global, self.history
-

build/lib/EDApy/__init__.py

@@ -0,0 +1,20 @@
+#!/usr/bin/env python
+# coding: utf-8
+
+# __init__.py
+
+import sys
+
+if sys.version_info[0] < 3:
+    raise Exception("Python version should be greater than 3")
+
+import warnings
+
+warnings.simplefilter(action='ignore', category=FutureWarning)
+from sklearn.cluster import KMeans
+
+from .optimization.multivariate.EDA_multivariate import EDAgbn as EDA_multivariate
+from .optimization.univariate.discrete import UMDAd as EDA_discrete
+from .optimization.univariate.continuous import UMDAc as EDA_continuous
+
+from .optimization.multivariate.__BayesianNetwork import print_structure as print_structure