Use print() function in both Python 2 and Python 3

airline/ann.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # The corresponding tutorial for this code was released EXCLUSIVELY as a bonus
 # If you want to learn about future bonuses, please sign up for my newsletter at:
 # https://lazyprogrammer.me
@@ -109,7 +110,7 @@ def fit(self, X, Y, activation=T.tanh, learning_rate=1e-3, mu=0.5, reg=0, epochs
                 c, p = train_op(Xbatch, Ybatch)
                 costs.append(c)
                 if (j+1) % print_period == 0:
-                    print "i:", i, "j:", j, "nb:", n_batches, "cost:", c
+                    print("i:", i, "j:", j, "nb:", n_batches, "cost:", c)
 
         if show_fig:
             plt.plot(costs)
@@ -156,16 +157,16 @@ def predict(self, X):
         X[:,d] = series[d:d+n]
     Y = series[D:D+n]
 
-    print "series length:", n
+    print("series length:", n)
     Xtrain = X[:n/2]
     Ytrain = Y[:n/2]
     Xtest = X[n/2:]
     Ytest = Y[n/2:]
 
     model = ANN([200])
     model.fit(Xtrain, Ytrain, activation=T.tanh)
-    print "train score:", model.score(Xtrain, Ytrain)
-    print "test score:", model.score(Xtest, Ytest)
+    print("train score:", model.score(Xtrain, Ytrain))
+    print("test score:", model.score(Xtest, Ytest))
 
     # plot the prediction with true values
     plt.plot(series)

airline/lr.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # The corresponding tutorial for this code was released EXCLUSIVELY as a bonus
 # If you want to learn about future bonuses, please sign up for my newsletter at:
 # https://lazyprogrammer.me
@@ -31,16 +32,16 @@
         X[:,d] = series[d:d+n]
     Y = series[D:D+n]
 
-    print "series length:", n
+    print("series length:", n)
     Xtrain = X[:n/2]
     Ytrain = Y[:n/2]
     Xtest = X[n/2:]
     Ytest = Y[n/2:]
 
     model = LinearRegression()
     model.fit(Xtrain, Ytrain)
-    print "train score:", model.score(Xtrain, Ytrain)
-    print "test score:", model.score(Xtest, Ytest)
+    print("train score:", model.score(Xtrain, Ytrain))
+    print("test score:", model.score(Xtest, Ytest))
 
     # plot the prediction with true values
     plt.plot(series)

airline/rnn.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # The corresponding tutorial for this code was released EXCLUSIVELY as a bonus
 # If you want to learn about future bonuses, please sign up for my newsletter at:
 # https://lazyprogrammer.me
@@ -87,7 +88,7 @@ def fit(self, X, Y, activation=T.tanh, learning_rate=1e-1, mu=0.5, reg=0, epochs
                 c = self.train_op(learning_rate, X[j], Y[j])
                 cost += c
             if i % 10 == 0:
-                print "i:", i, "cost:", cost, "time for epoch:", (datetime.now() - t0)
+                print("i:", i, "cost:", cost, "time for epoch:", (datetime.now() - t0))
             if (i+1) % 500 == 0:
                 learning_rate /= 10
             costs.append(cost)
@@ -141,7 +142,7 @@ def predict(self, X):
         X[:,d] = series[d:d+n]
     Y = series[D:D+n]
 
-    print "series length:", n
+    print("series length:", n)
     Xtrain = X[:n/2]
     Ytrain = Y[:n/2]
     Xtest = X[n/2:]
@@ -154,8 +155,8 @@ def predict(self, X):
 
     model = RNN([50])
     model.fit(Xtrain, Ytrain, activation=T.tanh)
-    print "train score:", model.score(Xtrain, Ytrain)
-    print "test score:", model.score(Xtest, Ytest)
+    print("train score:", model.score(Xtrain, Ytrain))
+    print("test score:", model.score(Xtest, Ytest))
 
     # plot the prediction with true values
     plt.plot(series)

bayesian_ml/1/nb.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # Naive Bayes with prior on mean and precision of Gaussian
 # mean | precision ~ N(0, c / precision)
 # precision ~ Gamma(a, b)
@@ -120,15 +121,15 @@ def plot_image(x, Q, title):
   Ytest = pd.read_csv('ytest.csv', header=None).as_matrix().flatten()
   model = NB()
   model.fit(Xtrain, Ytrain)
-  print "train accuracy:", model.score(Xtrain, Ytrain)
-  print "test accuracy:", model.score(Xtest, Ytest)
+  print("train accuracy:", model.score(Xtrain, Ytrain))
+  print("test accuracy:", model.score(Xtest, Ytest))
 
   # confusion matrix
   M = model.confusion_matrix(Xtest, Ytest)
-  print "confusion matrix:"
-  print M
-  print "N:", len(Ytest)
-  print "sum(M):", M.sum()
+  print("confusion matrix:")
+  print(M)
+  print("N:", len(Ytest))
+  print("sum(M):", M.sum())
 
   # plot 3 misclassified
   Q = pd.read_csv('Q.csv', header=None).as_matrix()

bayesian_ml/2/em.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # expectation-maximization for the model:
 # x(n) ~ N(Wz(n), sigma**2 I) (observed variables)
 # z(n) ~ N(0, I) (latent variables)
@@ -61,9 +62,9 @@ def loglikelihood(X, Z, W):
 plt.plot(costs)
 plt.show()
 
-print "actual W:", W0
-print "predicted W:", W
+print("actual W:", W0)
+print("predicted W:", W)
 
-print "log-likelihood given real W:", loglikelihood(X, Z, W0)
+print("log-likelihood given real W:", loglikelihood(X, Z, W0))
 
-print "log-likelihood found:", costs[-1]
+print("log-likelihood found:", costs[-1])

bayesian_ml/2/probit.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # probit regression
 
 import numpy as np
@@ -103,15 +104,15 @@ def plot_image(x, Q, title):
   Ytest = pd.read_csv('ytest.csv', header=None).as_matrix().flatten()
   model = ProbitRegression()
   model.fit(Xtrain, Ytrain)
-  print "train accuracy:", model.score(Xtrain, Ytrain)
-  print "test accuracy:", model.score(Xtest, Ytest)
+  print("train accuracy:", model.score(Xtrain, Ytrain))
+  print("test accuracy:", model.score(Xtest, Ytest))
 
   # confusion matrix
   M = model.confusion_matrix(Xtest, Ytest)
-  print "confusion matrix:"
-  print M
-  print "N:", len(Ytest)
-  print "sum(M):", M.sum()
+  print("confusion matrix:")
+  print(M)
+  print("N:", len(Ytest))
+  print("sum(M):", M.sum())
 
   # plot 3 misclassified
   Q = pd.read_csv('Q.csv', header=None).as_matrix()

bayesian_ml/3/run.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # variational-inference for linear regression
 # y(i) ~ N( x(i).dot(w), 1/lambda )
 # w ~ N( 0, diag(alpha_1, alpha_2, ..., alpha_D)^-1 )
@@ -81,7 +82,7 @@ def run(num=1, T=500):
   Y = pd.read_csv('y_set%s.csv' % num, header=None).as_matrix().flatten()
   Z = pd.read_csv('z_set%s.csv' % num, header=None).as_matrix().flatten()
   N, D = X.shape
-  print X.shape, Y.shape, Z.shape
+  print(X.shape, Y.shape, Z.shape)
 
   a0 = 1e-16
   b0 = 1e-16
@@ -129,16 +130,16 @@ def run(num=1, T=500):
     # update L
     L[t] = objective(X, Y, C, mu, a, b, e, f, a0, b0, e0, f0)
     if t % 20 == 0:
-      print "t:", t
+      print("t:", t)
       if num == 3:
-        print "L:", L[t]
+        print("L:", L[t])
 
   # plot 1/E[alpha]
   plt.plot(b/a)
   plt.show()
 
   # 1/E[lambda]
-  print "1/E[lambda]:", f/e
+  print("1/E[lambda]:", f/e)
 
   # plot L
   plt.plot(L)

bayesian_ml/4/emgmm.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # GMM using Expectation-Maximization
 
 import numpy as np
@@ -54,9 +55,9 @@ def gmm(X, K, max_iter=20, smoothing=1e-2):
   plt.scatter(X[:,0], X[:,1], c=R.argmax(axis=1))
   plt.show()
 
-  print "pi:", pi
-  print "means:", M
-  print "covariances:", C
+  print("pi:", pi)
+  print("means:", M)
+  print("covariances:", C)
   return R
 
 

bayesian_ml/4/npbgmm.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # GMM using Bayesian Nonparametric Clustering
 # Gaussian Mixture Model
 # Dirichlet Process
@@ -96,7 +97,7 @@ def gmm(X, T=500):
   observations_per_cluster = np.zeros((T, 6))
   for t in xrange(T):
     if t % 20 == 0:
-      print t
+      print(t)
     # 1) calculate phi[i,j]
     # Notes:
     # MANY new clusters can be made each iteration

bayesian_ml/4/vigmm.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # GMM using Variational Inference
 
 import numpy as np
@@ -199,7 +200,7 @@ def gmm(X, K, max_iter=100):
   plt.title("Costs")
   plt.show()
 
-  print "cluster assignments:\n", cluster_assignments
+  print("cluster assignments:\n", cluster_assignments)
   plt.scatter(X[:,0], X[:,1], c=cluster_assignments, s=100, alpha=0.7)
   plt.show()
 

best_fit_line.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 from pulp import *
 
 ### remove variable b because it is unconstrained
@@ -39,12 +40,12 @@
     prob += (a*x - y - c >= -z)
 
 status = prob.solve(GLPK(msg = 0))
-print "status:", LpStatus[status]
-print "values:"
-print "\ta:", value(a)
+print("status:", LpStatus[status])
+print("values:")
+print("\ta:", value(a))
 # print "\tb:", value(b)
-print "\tc:", value(c)
-print "\tz:", value(z)
+print("\tc:", value(c))
+print("\tz:", value(z))
 
 
 # extra part to plot everything

cnn_class/cifar.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # https://deeplearningcourses.com/c/deep-learning-convolutional-neural-networks-theano-tensorflow
 import os
 import numpy as np
@@ -43,7 +44,7 @@ def getImageData():
             im = Image.open("../large_files/cifar10/train/%s.png" % (i + 1))
             X[i] = image2array(im)
             if i % 1000 == 0:
-                print i
+                print(i)
         np.save(savedXpath, X.astype(np.uint8))
     else:
         X = np.load(savedXpath)
@@ -62,7 +63,7 @@ def getImageData():
             idx += 1
         Y[i] = label2idx[s]
         i += 1
-    print "done loading data"
+    print("done loading data")
     X, Y = shuffle(X, Y)
     return X[:30000], Y[:30000]
 
@@ -202,7 +203,7 @@ def fit(self, X, Y, lr=1e-4, mu=0.99, reg=1e-6, decay=0.99999, eps=1e-2, batch_s
                     c, p = cost_predict_op(Xvalid, Yvalid)
                     costs.append(c)
                     e = error_rate(Yvalid, p)
-                    print "i:", i, "j:", j, "nb:", n_batches, "cost:", c, "error rate:", e
+                    print("i:", i, "j:", j, "nb:", n_batches, "cost:", c, "error rate:", e)
 
         if show_fig:
             plt.plot(costs)

cnn_class2/tf_resnet_convblock_starter.py

@@ -31,5 +31,5 @@ def predict(self, X):
     conv_block.session = session
     session.run(init)
 
-    output = conv_block.predict(X):
+    output = conv_block.predict(X)
     print("output.shape:", output.shape)

hmm_class/generate_ht.py

@@ -35,9 +35,9 @@ def main():
         for n in range(50):
             sequence = generate_sequence(30)
             sequence = ''.join(symbol_map[s] for s in sequence)
-            print sequence
+            print(sequence)
             f.write("%s\n" % sequence)
 
 
 if __name__ == '__main__':
-    main()
+    main()

hmm_class/sites.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # https://deeplearningcourses.com/c/unsupervised-machine-learning-hidden-markov-models-in-python
 # https://udemy.com/unsupervised-machine-learning-hidden-markov-models-in-python
 # http://lazyprogrammer.me
@@ -19,14 +20,14 @@
     transitions[k] = v / row_sums[s]
 
 # initial state distribution
-print "initial state distribution:"
+print("initial state distribution:")
 for k, v in transitions.iteritems():
     s, e = k
     if s == '-1':
-        print e, v
+        print(e, v)
 
 # which page has the highest bounce?
 for k, v in transitions.iteritems():
     s, e = k
     if e == 'B':
-        print "bounce rate for %s: %s" % (s, v)
+        print("bounce rate for %s: %s" % (s, v))

numpy_class/dot_for.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # https://deeplearningcourses.com/c/deep-learning-prerequisites-the-numpy-stack-in-python
 # https://www.udemy.com/deep-learning-prerequisites-the-numpy-stack-in-python
 import numpy as np
@@ -23,4 +24,4 @@ def slow_dot_product(a, b):
 	a.dot(b)
 dt2 = datetime.now() - t0
 
-print "dt1 / dt2:", dt1.total_seconds() / dt2.total_seconds()
+print("dt1 / dt2:", dt1.total_seconds() / dt2.total_seconds())

numpy_class/manual_data_loading.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 # https://deeplearningcourses.com/c/deep-learning-prerequisites-the-numpy-stack-in-python
 # https://www.udemy.com/deep-learning-prerequisites-the-numpy-stack-in-python
 
@@ -16,4 +17,4 @@
   X.append(sample)
 
 X = np.array(X)
-print X
+print(X)