Revert number of iterations; Use numpy via wheels

.travis.yml

 # After changing this file, run it through http://lint.travis-ci.org/
 # Generates 4 builds, the ones for python3 will install external wheels
 language: python
-python:
-  - 2.6
-  - 2.7
 matrix:
   include:
+    - python: 2.6
+        - NUMPYSPEC===1.6.2
+    - python: 2.7
+        - NUMPYSPEC===1.8.0
     - python: 3.2
       env:
         - NUMPYSPEC===1.7.1

xbob/learn/misc/test_plda_trainer.py

@@ -362,8 +362,8 @@ def test_plda_EM_vs_Python():
 
   # Runs the PLDA trainer EM-steps (2 steps)
   # Defines base trainer and machine
-  t = PLDATrainer(1)
-  t_py = PythonPLDATrainer(max_iterations=1)
+  t = PLDATrainer(10)
+  t_py = PythonPLDATrainer(max_iterations=10)
   m = PLDABase(D,nf,ng)
   m_py = PLDABase(D,nf,ng)
 
@@ -375,8 +375,6 @@ def test_plda_EM_vs_Python():
   t.train(m, l)
   t_py.train(m_py, l)
   assert numpy.allclose(m.mu, m_py.mu)
-  print(m.f)
-  print(m_py.f)
   assert numpy.allclose(m.f, m_py.f)
   assert numpy.allclose(m.g, m_py.g)
   assert numpy.allclose(m.sigma, m_py.sigma)