Pythran: Crossing the Python Frontier [pdf]

#1

Post by serge-ss-paille » Tue, Apr 24, 2018, 6:48 AM UTC

Pythran: Crossing the Python Frontier [pdf]
computer.org

Re: Pythran: Crossing the Python Frontier [pdf]

#2

Post by accurrent » Tue, Apr 24, 2018, 8:46 AM UTC

Link to the actual software: https://github.com/serge-sans-paille/pythran

Re: Pythran: Crossing the Python Frontier [pdf]

#3

Post by eljost » Tue, Apr 24, 2018, 8:58 AM UTC

Interesting article but just skimming through it some things stand out immediately: 1.) The first snippet isn't even valid python code as floats don't have a shape attribute.

  s = 0. 
  n = s.shape

2.) The inline latex math isn't rendered properly.

Re: Pythran: Crossing the Python Frontier [pdf]

#4

Post by emmelaich » Tue, Apr 24, 2018, 9:00 AM UTC

accurrent wrote: post #2

Link to the actual software: https://github.com/serge-sans-paille/pythran

> "Pythran is an ahead of time compiler for a subset of the Python language, with a focus on scientific computing."

and

> a claimless python to c++ converter

Re: Pythran: Crossing the Python Frontier [pdf]

#5

Post by fermigier » Tue, Apr 24, 2018, 9:03 AM UTC

Talk from PyParis 2015: https://www.youtube.com/watch?v=Af8B30mXZ7E

Slides: https://fr.slideshare.net/PoleSystematicParisRegion/track-32...

Re: Pythran: Crossing the Python Frontier [pdf]

#6

Post by superdimwit » Tue, Apr 24, 2018, 10:12 AM UTC

Suspicious that they make brief mention of numba but don't make comparison to it in the following discussion or benchmarks..

Re: Pythran: Crossing the Python Frontier [pdf]

#7

Post by joshsyn » Tue, Apr 24, 2018, 10:36 AM UTC

Isn't this solved by julia? I think scientific community should use a more functional language rather than language like python tbh

Re: Pythran: Crossing the Python Frontier [pdf]

#8

Post by kristofferc » Tue, Apr 24, 2018, 10:37 AM UTC

eljost wrote: post #3

Interesting article but just skimming through it some things stand out immediately: 1.) The first snippet isn't even valid python code as floats don't have a shape attribute. s = 0. n = s.shape 2.) The inline latex math isn't rendered properly.

The first snippet also doesn't balance the parenthesis

    s += 100. * x[i + 1] - x[i] ** 2.) ** 2. + (1 - x[i]) ** 2

Re: Pythran: Crossing the Python Frontier [pdf]

#9

Post by rlayton2 » Tue, Apr 24, 2018, 10:42 AM UTC

eljost wrote: post #3

Interesting article but just skimming through it some things stand out immediately: 1.) The first snippet isn't even valid python code as floats don't have a shape attribute. s = 0. n = s.shape 2.) The inline latex math isn't rendered properly.

I believe the input should be a numpy array of floats which has a shape attribute

Re: Pythran: Crossing the Python Frontier [pdf]

#10

Post by montalbano » Tue, Apr 24, 2018, 10:43 AM UTC

joshsyn wrote: post #7

Isn't this solved by julia? I think scientific community should use a more functional language rather than language like python tbh

Though I'll still use Python for non-scientific programming, I've switched to Julia for all my scientific programming needs in my day job.