diff --git a/examples/plot_covariance_estimation.py b/examples/plot_covariance_estimation.py
index 2e0954957e64097f02ad10889dff273553256b1f..217397e22db08de851bc53f4926cb6ee0d6db0a7 100644
--- a/examples/plot_covariance_estimation.py
+++ b/examples/plot_covariance_estimation.py
@@ -40,7 +40,6 @@ negative_logliks = [-cov.fit(X_train, shrinkage=s).score(X_test) \
 
 ###############################################################################
 # Plot results
-pl.close('all')
 pl.loglog(shrinkages, negative_logliks)
 pl.xlabel('Shrinkage')
 pl.ylabel('Negative log-likelihood')
diff --git a/examples/plot_ica_vs_pca.py b/examples/plot_ica_vs_pca.py
index 76509a3cfa71b7558b7d74c76c593218911d1492..4787cecda4f80ea843e6a1f2d4c39dda96d18bd6 100644
--- a/examples/plot_ica_vs_pca.py
+++ b/examples/plot_ica_vs_pca.py
@@ -77,7 +77,6 @@ def plot_samples(S, axis_list=None):
     pl.xlabel('$x$')
     pl.ylabel('$y$')
 
-pl.close('all')
 pl.subplot(2, 2, 1)
 plot_samples(S / S.std())
 pl.title('True Independant Sources')
diff --git a/examples/plot_permutation_test_for_classification.py b/examples/plot_permutation_test_for_classification.py
index 7f66b98e27f2ce058c9463cfb4890026a10b612e..d41eb0c3c78d24704a14ae59fc7bdfe95f2258eb 100644
--- a/examples/plot_permutation_test_for_classification.py
+++ b/examples/plot_permutation_test_for_classification.py
@@ -17,6 +17,7 @@ obtained in the first place.
 print __doc__
 
 import numpy as np
+import pylab as pl
 
 from scikits.learn.svm import SVC
 from scikits.learn.cross_val import StratifiedKFold, permutation_test_score
@@ -49,15 +50,13 @@ print "Classification score %s (pvalue : %s)" % (score, pvalue)
 
 ###############################################################################
 # View histogram of permutation scores
-import pylab as pl
-pl.close('all')
 pl.hist(permutation_scores, label='Permutation scores')
 ylim = pl.ylim()
 pl.vlines(score, ylim[0], ylim[1], linestyle='--',
-                        color='g', linewidth=3, label='Classification Score'
-                                                      ' (pvalue %s)' % pvalue)
+          color='g', linewidth=3, label='Classification Score'
+          ' (pvalue %s)' % pvalue)
 pl.vlines(1.0 / n_classes, ylim[0], ylim[1], linestyle='--',
-                        color='k', linewidth=3, label='Luck')
+          color='k', linewidth=3, label='Luck')
 pl.ylim(ylim)
 pl.legend()
 pl.xlabel('Score')