First PR

[terrytangyuan]

One test failed on my local mac pc. Let's see what Travis will return.

======================================================================
ERROR: test_iris (test.metric_learn_test.TestSDML)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/yuantang/Desktop/metric_learn/test/metric_learn_test.py", line 79, in test_iris
    sdml = SDML().fit(self.iris_points, graph)
  File "/Users/yuantang/Desktop/metric_learn/metric_learn/sdml.py", line 52, in fit
    self.M, _ = graph_lasso(emp_cov, self.sparsity_param, verbose=verbose)
  File "/Library/Python/2.7/site-packages/sklearn/covariance/graph_lasso_.py", line 255, in graph_lasso
    raise e
FloatingPointError: Non SPD result: the system is too ill-conditioned for this solver. The system is too ill-conditioned for this solver

[terrytangyuan]

@perimosocordiae Travis succeeded. I don't know why it became ill-conditioned when it comes to testing on my local pc.

[perimosocordiae]

Thanks, merged!