bcapps VS PolygonDistances

Just for fun, I created maxdierolls.csv in https://github.com/barrycarter/bcapps/tree/master/ONEOFF/ that shows the results for 1 <= k <= n <= 100

To be fair, I actually do this in https://github.com/barrycarter/bcapps/tree/master/bc-gcdist.pl because it's useful in determining Internet ping times. Roughly speaking, Internet packets rarely travel faster than 0.5c (though I have seen as high as 0.6c)

I never really got anywhere on this, but my work is at https://github.com/barrycarter/bcapps/tree/master/bc-calc-risk-odds.m

The exact chance is 34382839/81839428 or about 42.013%. I used Mathematica and recursion to solve this: https://github.com/barrycarter/bcapps/blob/master/REDDIT/repeat-cards.m

Feel free to take a look at https://github.com/barrycarter/bcapps/tree/master/README-projects.txt but here's a more sinister and useful suggestion that doesn't really answer your question.

Try graph1.csv in https://github.com/barrycarter/bcapps/blob/master/ONEOFF/

My https://github.com/barrycarter/bcapps/tree/master/FINANCES/bc-parse-bank-email.pl probably won't help, but is an example of how you can parse different types of documents (emails from various banks) using a single script

https://github.com/barrycarter/bcapps/tree/master/bc-getmail.pl checks all my email accounts in a neverending loop, handling cases where getmail gets stuck

At one point, I cataloged some PBS strips at https://github.com/barrycarter/bcapps/tree/master/METAWIKI/pbs.txt -- you might find something useful there though it's nowhere near complete

The fastest thing I've found so far is to rasterize the data, project the points into 3D, create a KDTree, compute the straight line distances, and convert those to spherical distances, all using numpy with python (not SQL), but I wonder if I'm missing something. My work so far: https://github.com/barrycarter/PolygonDistances/

Thought it would be fun to do that for other countries that have territories spread across the globe like the UK and France. Or things like Europe or the Commomwealth Realms or something. Some code so far: https://github.com/barrycarter/PolygonDistances/

https://stackoverflow.com/questions/54133042/difference-between-numpys-vectorize-and-its-automatic-vectorization may help. Instead of speeding up np.vectorize, don't use it at all. Instead, do parallelizable math on ndarrays, and Python/numpy will automatically do the math in parallel. As an example run this fast code (also at https://github.com/barrycarter/PolygonDistances as sample2.py):

If you'd be interested in learning about GIS with Python with someone who is learning it themselves, ping me: https://github.com/barrycarter/PolygonDistances