dklend122

8 ms·

31.
Data Science in Julia for Hackers(datasciencejuliahackers.com)
240 points by dklend122 6y ago|51 comments
32.
JuliaSymbolics Roadmap: A Modern Computer Algebra System for a Modern Language(juliasymbolics.org)
3 points by dklend122 6y ago|0 comments
33.
Floops.jl: unified system for safe threaded, distributed and GPU loops in Julia(github.com)
4 points by dklend122 6y ago|0 comments
34.
by dklend122 6y ago
So you make the type lattice user extendable? Or different layers hardcoded? If the former, how do you get code reuse? What do you mean by expanded lattice? I thought the type lattice was fixed. What would something general look like ? Than
35.
by dklend122 6y ago
Interesting. How do you reconcile this: "Julia's type system is designed for semantics specification, not proofs, so it's not really what you'd want to use there." With "think the most promising way to go is ba
36.
by dklend122 6y ago
Yea. Hopefully interfaces, which have been mentioned by the core team as being desired , will help.
37.
by dklend122 6y ago
To your second point, it will throw method errors at compile time
38.
by dklend122 6y ago
Ah that's true in spirit, but it wasn't entirely false.
39.
by dklend122 6y ago
Also, the first run of "@time" includes compilation. So for an accurate runtime, you'd want to run it twice and use the second timing.
40.
by dklend122 6y ago
Which version of Julia did you use? @view didn't actually speed up code until 1.5, where they got stack allocated.
41.
by dklend122 6y ago
Ty
42.
by dklend122 6y ago
There are many compelling reasons. Doesn't seem that way if Julia is just construed as a faster python with some Matlab. Goes far beyond that
43.
by dklend122 6y ago
Mathematical programming in Julia is just more natural and integrated with the language and type system. From dot broadcasting, to multiple dispatch, to parametric types (if you want ), multidimensional array comprehensions. No worrying ab
44.
by dklend122 6y ago
It wasn't too long ago that the reverse happened. People are excited about Julia, because frankly it's amazing and a breath of fresh air coming from python. Also, it's really now starting to pick up steam. Curious though, how
45.
by dklend122 6y ago
The core team has plans to add optional static typing to julia..won't be there for a year or two though probably
46.
by dklend122 6y ago
I understand CPU is slower, but it's not going anywhere, and matmul is just a benchmark. It's extremely impressive that a high level (using index notation for multiple backends) Julia library can compete with hand tuned kernels. A
47.
by dklend122 6y ago
No, the parent comment is correct. Tulllio.jl has been getting up to openBLAS levels of performance Julia also is on the cusp of support for tensor cores, mixed precision multiplication and bfloat16 codegen. New PR just merged last week
48.
by dklend122 6y ago
There are many areas in which Julia's ecosystem is far ahead of python's, today. Standard ML will catch up eventually
49.
Using type domain information in Julia(ericphanson.com)
8 points by dklend122 6y ago|0 comments
50.
by dklend122 6y ago
Download metrics show usage doubling every year. As I said above, There's no coordinated effort to do this. These are all independent pieces that were written at different times by different people.
51.
by dklend122 6y ago
Personally, I much prefer end to braces. They have the benefit also of being usable for indexing and array expressions
52.
by dklend122 6y ago
This was part of a tutorial series for a data science program The "unreasonable effectiveness" article wasn't, to my knowledge, by a community member. It's normal for magazines to have enthusiastic headlines. You're
53.
by dklend122 6y ago
I think it's parsed specially, but I could be wrong. Definitely no runtime check
54.
by dklend122 6y ago
Julia already has an interpreter, and the core team has plans for tiered compilation. Also the JIT compile time can be minimized to the point where it's a very minor thing. There's also smarter and slimmer static compile, which is
55.
by dklend122 6y ago
There are several: https://arstechnica.com/science/2020/10/the-unreasonable-eff... Also comes with some, but relatively light trade-offs for what's gained, IMO. The worst of trade-offs, compile overheads
56.
by dklend122 6y ago
"the", graph plotting library? There are several and gadfly isn't the most popular. Compile times have dropped precipitously, around 5 sec on master for plots.jl. less than one second if you use the package compiler. This doe
57.
by dklend122 6y ago
There's no coordinated effort to do this. These are all independent pieces that were written at different times by different people. Where do you see the arrogance in the OP? It's showing exactly what's being claimed
58.
by dklend122 6y ago
No, that's not true. I specifically mentioned features that can easily be made zero cost on both CPU and GPU targets. In fact, those aren't special features and there is no other subset. Those are the core abstractions on which e
59.
by dklend122 6y ago
Where do you you see these mythical quick strides? Do you have data to back up this claim ?
60.
by dklend122 6y ago
No. You're confusing static dispatch with static typing. Julia can do both static and dynamic dispatch. The latter works with function barriers works to maintain codebases that are impossible to achieve with just one or the other. Ther

More ›