Changed test to match python API and update README to match output

[alvarna]

Hi,

Thanks for the tool ! Looks like the python test was using an older API. I updated it to pass the projection set to the count function and updated the README to match the output, which is represented differently (but numerically the same).

[msoos]

Ah nice! Thank you! I have actually pushed a change to the README... however, I would very much appreciate the fix to the test_pyapproxmc.py. Can you please change this PR to only contain the diff to the test and then I will approve & merge?

Thanks so much again,

Mate

[alvarna]

The update in this PR was different from #37 - This updates the example in the README in the root of the repo from
Approximate count is: 88*2**13
to:
Approximate count is: 11*2**16
which matches the output I get currently. These are numerically the same values, but it looks like the no. of hashes has increased, and the no. of cells has correspondingly reduced. Similarly, for the second example, the output from the current version is Approximate count is: 7*2**6 (instead of 56*2**3).

I can update the example in the python README to match this as well, if you wish. Or if you prefer to leave it with the old output, that is fine with me as well. Let me know how you would like to proceed.

[msoos]

Thanks this is perfect! Thanks for updating it to the currently displayed one :) Really nice! Thanks again!