datanecdote

22 ms·

1.
by datanecdote 6y ago
As the document I linked to says, Jax autograd supports custom data types and custom gradients. It’s honestly exhausting arguing with all you Julia boosters. You can down vote me to hell, I don’t care. I’m done engaging with this community.
2.
by datanecdote 6y ago
Try reading the docs before making sweeping negative comments about what a piece of software can and cannot do. https://jax.readthedocs.io/en/latest/notebooks/autodiff_cook...
3.
by datanecdote 6y ago
It is true that Jax cannot differentiate through C code. But it can differentiate through python code that was written to accept Numpy.
4.
by datanecdote 6y ago
For Jax I believe this is false. Jax is composable. In fact it’s a core design goal. Jax arrays implement the Numpy API. I routinely drop Jax arrays into other python libraries designed for Numpy. It works quite well. It’s not effortless 10
5.
by datanecdote 6y ago
How does Jax lose composability or introspection?
6.
by datanecdote 6y ago
Lengthy, nuanced discussion about benchmarking between Turing devs and Stan devs.
7.
by datanecdote 6y ago
The right benchmark is Stan https://github.com/TuringLang/TuringExamples/pull/25
8.
by datanecdote 6y ago
When I look at google trends or redmonk rankings, Julia appears stable, not accelerating.
9.
by datanecdote 6y ago
Why minizinc instead of Google OR? Seems like Google OR best minizinc at their own contest? https://www.minizinc.org/challenge2020/results2020.html Is it more customizable? Or expressive (in terms of modeling DSL)?
10.
by datanecdote 6y ago
Thanks Viral. To be clear, I’m a python user who’s cheering for Julia, because I live the problems of python and do see the potential of Julia as a better path. But unfortunately I’m not prepared to be the early adopter (at least in my day
11.
by datanecdote 6y ago
> 2. I don't entirely follow this point. Perhaps using PyArrow's parser would be faster than what is timed here, but is that what the typical Python data science user would do? I am a Python data science user. If data gets big
12.
by datanecdote 6y ago
@Sukera Fair, but, if I break up all the loops and if statements into functions, those functions still have “end”s
13.
by datanecdote 6y ago
That was mostly meant as a joke, thus the “;-)” I don’t really care much about syntax choices, but my small complaint about “end” is that it takes up a line which reduces the amount of business-logic code I can fit on one screen, especially
14.
by datanecdote 6y ago
Thanks. I watched the JuliaCon state of Julia presentation. As I wrote in my original post, I appreciate the investments the Julia core developers are making, that have improved but not eliminated this problem. I wish them luck.
15.
by datanecdote 6y ago
Preach, brother. I’m cautiously optimistic that JAX (or something like JAX) can save the python programming language from stagnation by essentially building a feature-complete reimplementation of the language with JIT and autograd baked int
16.
by datanecdote 6y ago
Let me second GP’s sentiment. I find Julia really slow for my purposes. I don’t know his reasoning, but I will explain mine. None of this is surprising and is oft discussed. Julia (at least by default) is constantly recompiling everything.
17.
by datanecdote 6y ago
FWIW I agree with you. I’ve always found cython easier than Numba. And more performant. I think Numba has a lot of potential and will improve as they fill out remaining language coverage and finalize the API. The idea of a LLVM JIT compiler
18.
by datanecdote 6y ago
In the uncommon event I need to write a loop from scratch, and I need it to be really fast, I just rewrite that one jupyter cell in cython or numba. But that is a small piece of my codebase. I agree that Julia code is aesthetically superior
19.
by datanecdote 6y ago
Honest question from a heavy python user who would switch if it made sense: Are there any comprehensive benchmarks that show Julia outperforming Pandas or PyTorch or SciKit? Obviously pure Python is terrible. But the library algorithms writ
20.
by datanecdote 6y ago
I’m fascinated by Julia and have test driven it before but it didn’t click for me. Maybe I was doing it wrong and/or the ecosystem has matured since I last looked. I guess I generally do like the pythonic paradigm of an interpreted glu
21.
by datanecdote 6y ago
Thanks for the pointers, those crates seem great. The flaky multithreading libs are my least favorite part of python, and rust’s strength in this area seems very appealing.
22.
by datanecdote 6y ago
I deal with a lot of ragged data that is hard to vectorize, and currently write cython kernels when the inner loops take too long. Sounds like Rust might be faster than cython? Thanks for the feedback.
23.
by datanecdote 6y ago
I am in similar boat. Python centric data scientist. Very tempted to try to learn Rust so I can accelerate certain ETL tasks. Question for Rust experts: On what ETL tasks would you expect Rust to outperform Numpy, Numba, and Cython? What ar
24.
by datanecdote 6y ago
Worth keeping in mind the principal agent problem literature. Selling equity in a company is vulnerable to “lemon” problems. Startups with product market fit want to minimize dilution and keep executing before raising at a higher valuation