3 ms·

Reddit AMA: We are the creators of the Julia programming language

11 points by yarapavan 6y ago

tetris11 6y ago
Yawn. Also, since when was Julia so popular with the average redditor such that this AMA could be found on the frontpage? I smell a paid promotion.
- KenoFischer 6y ago
  I think if I had spare money lying around to bribe reddit moderators, I'd rather buy myself a nice server, or lots of steak dinners, or a new oscilloscope. Honestly, I can think of lots of better uses of my money than that.
- ViralBShah 6y ago
  You smell wrong.
smabie 6y ago
I've recently been using Julia to analyze around 180gb of CSV's and it's been really nice. The problem with Python (besides the sucky language) is that it's so slow you have to work pretty hard to make sure all your operations are calling into some C lib. Trying to call a python function over 5m pandas rows? Forget about it, never going to happen. Anything that requires an if statement, never going to happen. Often you can express what you're trying to do in terms of vectorized numpy/pandas operations, but it takes some work. Julia simply doesn't have this problem and it's a game changer. By and large, everything is reasonably fast and the compiler does a good job. Julia has made scientific computing so much more accessible than Python ever could. Anyone can publish a library and even if the code is rough, it'll usually be fast enough to be useful. With Python, unless you're a professional software engineer and willing to write in C, forget about sharing a new model or numerical technique with anyone else. I think the lack of classes and strong support for function polymorphism has made Julia code pretty easy to write, easy to read, and very powerful. In Python each library is its own little world and you can't really mix and match them in a meaningful way. In Julia you can leverage multiple libraries in new and unexpected ways. Perhaps the best part of Julia is the broadcasting operator (.) that allows any function to be easily mapped onto a collection. It's almost like using an array language like k or j! Also, macros. And with the seamless Python interop, Julia should be your goto pick for any scientific computing project. It's even starting to replace C and Fortran, which I think is pretty exciting. The future couldn't be brighter for Julia!

11 points by yarapavan 6y ago

Yawn. Also, since when was Julia so popular with the average redditor such that this AMA could be found on the frontpage? I smell a paid promotion.

KenoFischer 6y ago
I think if I had spare money lying around to bribe reddit moderators, I'd rather buy myself a nice server, or lots of steak dinners, or a new oscilloscope. Honestly, I can think of lots of better uses of my money than that.
ViralBShah 6y ago
You smell wrong.

smabie 6y ago

I've recently been using Julia to analyze around 180gb of CSV's and it's been really nice. The problem with Python (besides the sucky language) is that it's so slow you have to work pretty hard to make sure all your operations are calling into some C lib. Trying to call a python function over 5m pandas rows? Forget about it, never going to happen. Anything that requires an if statement, never going to happen. Often you can express what you're trying to do in terms of vectorized numpy/pandas operations, but it takes some work. Julia simply doesn't have this problem and it's a game changer. By and large, everything is reasonably fast and the compiler does a good job. Julia has made scientific computing so much more accessible than Python ever could. Anyone can publish a library and even if the code is rough, it'll usually be fast enough to be useful. With Python, unless you're a professional software engineer and willing to write in C, forget about sharing a new model or numerical technique with anyone else. I think the lack of classes and strong support for function polymorphism has made Julia code pretty easy to write, easy to read, and very powerful. In Python each library is its own little world and you can't really mix and match them in a meaningful way. In Julia you can leverage multiple libraries in new and unexpected ways. Perhaps the best part of Julia is the broadcasting operator (.) that allows any function to be easily mapped onto a collection. It's almost like using an array language like k or j! Also, macros. And with the seamless Python interop, Julia should be your goto pick for any scientific computing project. It's even starting to replace C and Fortran, which I think is pretty exciting. The future couldn't be brighter for Julia!