joehl

13 ms·

1. 1.▲
   GreeNsort: Simple Symmetric Sustainable Sorting(greensort.org)

   1 points by joehl 3y ago|0 comments
2. 2.▲
   by joehl 6y ago

   If your data set is biased for the purpose of answering your question, then having more data doesn't make things better. In this case 'big data' just becomes 'big bias'. Go to wikipedia and read on Simpsons Paradox.
3. 3.▲
   by joehl 6y ago

   In causal modelling thou shallst not control by consequences of the causal treatment (the causal treatment here is being born black). Read your Rubin&Rosenbaum.
4. 4.▲
   by joehl 7y ago

   Providing Julia Code for in-memory sorting, i.e. a binary mergesort with parallel branches and parallel merge plus a specialized sort for instable inplace sorting (better Quicksort, parallel branches only) is not a huge step from the code a
5. 5.▲
   by joehl 7y ago

   OK, I found out myself: Julia can mutate a single temp buffer across all threads, and implementing Sedgewicks nocopy sort (with ping-pong merge) is faster, particularly when passing down preallocated buffer memory to the single threaded
6. 6.▲
   by joehl 7y ago

   Maybe I misunderstand, but: In R allocating and copying huge return vectors for small changes on input vectors ruins performance. My understanding of Julia was that one advantage over R is the possiblity to mutate input vectors in place. Re