tadlan

6 ms·

1.
by tadlan 11y ago
Gotcha!
2.
by tadlan 11y ago
Thanks. I'm not familiar with dynamic linking. What will that do?
3.
by tadlan 11y ago
Any specific plans for python support?
4.
by tadlan 11y ago
Or julia. Which is 1000th better than js
5.
by tadlan 11y ago
It will eventually, iirc
6.
by tadlan 11y ago
Question please: Even though there is stated desire to support python etc vm, this will be hampered by the need for the client to download the large VM vs native code that can be distributed itself, correct?
7.
by tadlan 11y ago
Have you tried dask?
8.
by tadlan 11y ago
What about dynd? Iirc doesn't use iteration st all.
9.
by tadlan 11y ago
Woah!!! Cool
10.
by tadlan 11y ago
Make sure you use the devect macro or parallel accelerator if you're going to use vector iced code :)
11.
by tadlan 11y ago
It's not a would and gpod. Numba works like that right now, has better interoperability with python and less context switching. Numba is I easily installed on all systems through conda package manager and is easily used with just a fun
12.
by tadlan 11y ago
Actually, Numba array allocation in nopython mode is allowed now. What other numerical features are missing?
13.
by tadlan 11y ago
Numba obviate these issues
14.
by tadlan 11y ago
What would one need to use c or c++ to write a julia package? It's as fast as native code so no need to use multiple languages.
15.
by tadlan 11y ago
Numba would still be faster. It would fuse away any intermediates in the code and remove any Overhead to the compiled code. also have the option of devecting to loops. both of which are generally faster than vectorized jumpy code.
16.
by tadlan 11y ago
Numbs compiles entire functions on and allows array expressions with allocation and loop fusion. I don't see the problem
17.
by tadlan 11y ago
Use numba to compile python loops or array expressions to fast llvm, and problem solved. I'm sticking with python.
18.
by tadlan 11y ago
That's just the data structure. Fitting models would be much tougher.
19.
by tadlan 11y ago
Will julia ever be popular for general computing and Web stuff? It's s till fast for non scientific use cases, has multi threading (in master) and is getting go like channels.
20.
by tadlan 11y ago
Julia!!!!! The answer to all these problems, both for general and science programming.
21.
by tadlan 11y ago
Dask has that, and scikit learn is moving that way also. It even beats spark for out of core work on a single machine
22.
by tadlan 11y ago
There's a dev branch that has been actively developed
23.
by tadlan 11y ago
Numba has list, Named tuples, and is working on define your own jit classes.
24.
by tadlan 11y ago
Numba has loop fusion, runtime, and soon multithreading parallel vectorized functions. How is that c like?
25.
by tadlan 11y ago
Check out Julia ' s Cxx.jl. it's probably the best c++ FFI in any language. You can even evaluation and call cpp code on the fly from the repl
26.
by tadlan 11y ago
What was the diet?
27.
by tadlan 11y ago
That is on the roadmap
28.
by tadlan 11y ago
Or use pymc 3 for the bayes equivalent. It's Better anyways. Yes one can also use rpy2. I think glmm is a gsoc project or an upcoming pr but not 100% sure.
29.
by tadlan 12y ago
What are you implying?
30.
by tadlan 12y ago
What is second generation numba? And any plans to branch numba out of pure numeric application s?

More ›