Python deployed example fix

[jklaise]

Updated the Python deployed model example to respect the API changes for both Skater and Lime since the example was written. Also replaced the binary trained model file with a new one as the old one throws an error when trying to load via joblib.

[coveralls]

Coverage remained the same at 62.572% when pulling 0f5b7f643aaecff602839f74c1ed83a855da27ca on jklaise:python-example-fix into 79e85bc2e28a8379e91a3ff07a690dace31e3289 on datascienceinc:master.

[pramitchoudhary]

This is great @jklaise. Did you have to modify an example notebook? or do you have one which you think could be added?

[jklaise]

@pramitchoudhary I had to modify the existing example notebook because the example doesn't work as it's based on the (presumably) old API. As I've mentioned in #261, this is just one place where it doesn't work, there are multiple places in the docs and the examples using the old API (calling Interpretation() without arguments).

[pramitchoudhary]

Apologies for the delay in merging the changes @jklaise.

[sohamxi]

Is the pull request complete?