How to handle CRLs correctly in long-term electronic signatures?

What is the best way to handle sparse data on scikit-learn?

• I am writting an instance reduction module on scikit-learn. It is working, but when compiling/testing it gives following error: =================================================ERROR: sklearn.tests.test_common.test_estimators_sparse_data ---------------------------------------------------------------------- Traceback (most recent call last):   File "/usr/lib/python2.7/dist-packages/nose/http://case.py", line 197, in runTest     self.test(*self.arg)   File "/home/dayvid/workspace/scikit-learn/sklearn/tests/http://test_common.py", line 145, in test_estimators_sparse_data     raise exc ValueError: data type not understood Here is one of the class signatures: class CondensedNearestNeighbors(BaseEstimator, ClassifierMixin, InstanceReductionMixin) If I use scipy.sparse.issparse(X), it says that it doesn't fail gracefully. Please, help, I have over 14 algorithms (some of them algorithms of this year) and I'm really excited about contributing to sklearn. Thanks

• Answer:

  Upgrade to 0.14 and use k-means!(for speed) PS: It helps more if you pick the domain of your problem and pick the tools for that domain e.g. or-tools for OR problems,  cvxopt if you want to do convex optimisation ....rather than diving in to split wide open esp as a beginner. sklearn random forests,for instance, do not work on sparse inputs . It d slap you with "ValueError: array is too big." if you do that and so will Gmm, [1]Been there , did the latter and hated the burden.