python3 fix

wisard.py

@@ -28,8 +28,8 @@ def printProgressBar(label,time,etime,basecolor, cursorcolor, linecnt,progress,s
             str += color.END + cursorcolor + u'\u2588' + color.END + basecolor
         else:
             str += u'\u2591'
-    str += color.END + '| ' + "{:>3}".format(int(progress * 100.0)) + ' % ' + color.YELLOWBLACK + ' ' + etime + ' ' + color.WHITEBLACK + time + ' ' + color.END
-    sys.stdout.write("\r%s" % str.encode('utf-8'))
+    str += color.END + '| ' + "{:>3}".format(int(progress * 100.0)) + ' % ' + color.BLUE + ' ' + etime + ' ' + color.BLUE + time + ' ' + color.END
+    sys.stdout.write("\r%s" % str)
     sys.stdout.flush()
     return progress
 
@@ -42,21 +42,21 @@ def compTime(deltatime,progress):
     tme = "{:0>2}:{:0>2}:{:02.0f}".format(int(hourse),int(minutese),secondse)
     return tm,tme
 
-def decide_onebyone((clf,data)):
+def decide_onebyone(clf,data):
     return [clf.wiznet_[cl].Classify(data, clf.ranges_, clf.offsets_, clf.notics) for cl in clf.classes_]
 
-def decide_onebyone_noscale((clf,data)):
+def decide_onebyone_noscale(clf,data):
     return [clf.wiznet_[cl].ClassifyNoScale(data, clf.notics) for cl in clf.classes_]
 
-def train_onebyone((clf,X,y)):
+def train_onebyone(clf,X,y):
     for i,data in enumerate(X):
         clf.wiznet_[clf.classes_[y[i]]].Train(data,clf.ranges_,clf.offsets_, clf.notics)
 
-def train_onebyone_noscale((clf,X,y)):
+def train_onebyone_noscale(clf,X,y):
     for i,data in enumerate(X):
         clf.wiznet_[clf.classes_[y[i]]].TrainNoScale(data, self.notics)
 
-def decide_onebyone_b((clf,data)):
+def decide_onebyone_b(clf,data):
     b = clf.b_def
     confidence = 0.0
     res_disc_list = [clf.wiznet_[cl].Response(data,clf.ranges_,clf.offsets_, clf.notics) for cl in clf.classes_]
@@ -76,7 +76,7 @@ def decide_onebyone_b((clf,data)):
         result = np.array(result_partial)/result_sum
     return result
 
-def decide_onebyone_b_noscale((clf,data)):
+def decide_onebyone_b_noscale(clf,data):
     b = clf.b_def
     confidence = 0.0
     res_disc_list = [clf.wiznet_[cl].ResponseNoScale(data, clf.notics) for cl in clf.classes_]
@@ -402,7 +402,7 @@ def decision_function_par_b_noscale(self,X):    # parallel version (no debug wit
     def decision_function_seq_b(self,X):    # sequential version (no debug with bleaching)
         D = np.empty(shape=[0, len(self.classes_)])
         for data in X:
-            res = decide_onebyone_b((self,data))  # classify with bleaching (Work in progress)
+            res = decide_onebyone_b(self,data)  # classify with bleaching (Work in progress)
             D = np.append(D, [res],axis=0)
         return D
 
@@ -436,7 +436,7 @@ def decision_function_seq_b_debug(self,X):
         self.starttm_ = time.time()
         self.progress_ = 0.01
         for data in X:
-            res = decide_onebyone_b((self,data))  # classify with bleaching (Work in progress)
+            res = decide_onebyone_b(self,data)  # classify with bleaching (Work in progress)
             D = np.append(D, [res],axis=0)
             cnt += 1
             tm,tme = compTime(time.time()-self.starttm_,self.progress_)
@@ -450,7 +450,7 @@ def decision_function_seq_b_debug_noscale(self,X):
         self.starttm_ = time.time()
         self.progress_ = 0.01
         for data in X:
-            res = decide_onebyone_b_noscale((self,data))  # classify with bleaching (Work in progress)
+            res = decide_onebyone_b_noscale(self,data)  # classify with bleaching (Work in progress)
             D = np.append(D, [res],axis=0)
             cnt += 1
             tm,tme = compTime(time.time()-self.starttm_,self.progress_)

wisard_wrapper.pyx

@@ -32,7 +32,7 @@ cdef extern from "Discriminator.h" namespace "wnn":
 cdef class PyDiscriminator:
     cdef Discriminator *thisptr      # hold a C++ instance which we're wrapping
     def __cinit__(self, bits, size, maptype="random"):
-        self.thisptr = new Discriminator(bits, size, maptype)
+        self.thisptr = new Discriminator(bits, size, maptype.encode('utf-8'))
     def __dealloc__(self):
         del self.thisptr
     def getNBits(self):