EinsteinPy

19 views
Skip to first unread message

Philip Thrift

unread,
May 4, 2020, 8:02:23 PM5/4/20
to Everything List

Latest release:

https://twitter.com/EinsteinPy/status/1257452756413165568

@philipthrift

Russell Standish

unread,
May 4, 2020, 9:26:41 PM5/4/20
to Everything List
Yes - Python is the duck's nuts right now. I want to switch TCL out for
Python in my EcoLab platform, which I use for ALife research. It'll
probably have a to wait a few years until I "retire", though, and I
hope that the scientific community hasn't jumped on another
bandwagon by that time :).

On Mon, May 04, 2020 at 05:02:23PM -0700, Philip Thrift wrote:
>
>
> Latest release:
>
> https://twitter.com/EinsteinPy/status/1257452756413165568
>
> @philipthrift
>
> --
> You received this message because you are subscribed to the Google Groups "Everything List" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/e97522ce-5c96-4459-af0c-fb790517e4d6%40googlegroups.com.


--

----------------------------------------------------------------------------
Dr Russell Standish Phone 0425 253119 (mobile)
Principal, High Performance Coders hpc...@hpcoders.com.au
http://www.hpcoders.com.au
----------------------------------------------------------------------------

smitra

unread,
May 5, 2020, 4:47:22 AM5/5/20
to everyth...@googlegroups.com
Isn't Python too slow for such simulation work?
Saibal

Philip Thrift

unread,
May 5, 2020, 6:35:01 AM5/5/20
to Everything List

A beginners guide to using Python for performance computing

JIT fast! Supercharge tensor processing in Python with JIT compilation

...

@philipthrift



On Tuesday, May 5, 2020 at 3:47:22 AM UTC-5, smitra wrote:
Isn't Python too slow for such simulation work?

On 05-05-2020 03:25, Russell Standish wrote:
> Yes - Python is the duck's nuts right now. I want to switch TCL out for
> Python in my EcoLab platform, which I use for ALife research. It'll
> probably have a to wait a few years until I "retire", though, and I
> hope that the scientific community hasn't jumped on another
> bandwagon by that time :).
>
> On Mon, May 04, 2020 at 05:02:23PM -0700, Philip Thrift wrote:
>>
>>
>> Latest release:
>>
>> https://twitter.com/EinsteinPy/status/1257452756413165568
>>
>> @philipthrift
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "Everything List" group.
>> To unsubscribe from this group and stop receiving emails from it, send

ronaldheld

unread,
May 5, 2020, 2:30:27 PM5/5/20
to Everything List
Does anyone use this?   IMO, Pytjon is slower than my F77/90 executables, but I have never tried a direct comparison.  IMO, it is Pytjon-mania.
     Ronald

Russell Standish

unread,
May 5, 2020, 10:25:46 PM5/5/20
to everyth...@googlegroups.com
On Tue, May 05, 2020 at 10:47:15AM +0200, smitra wrote:
> Isn't Python too slow for such simulation work?

The heavy lifting is done in C++, with full HPC support (OpenMP and
MPI). EcoLab naturally embeds an interpreter, which is currently TCL,
enabling rapid setting up of different experiements. Thinki of it like
Tensorflow, not "everything done in Python".

Philip Thrift

unread,
May 6, 2020, 2:23:29 AM5/6/20
to Everything List

There is typically support for weaving other coding languages into Python code:

There are three basic ways to use weave. The weave.inline() function executes C code directly within Python, and weave.blitz() translates Python NumPy expressions to C++ for fast execution. blitz() was the original reason weave was built. For those interested in building extension libraries, the ext_tools module provides classes for building extension modules within Python.



@phulipthrift

Telmo Menezes

unread,
May 6, 2020, 6:14:39 AM5/6/20
to Everything List
I am also very fond of Cython. It is a superset of Python that complies to C, and it is easy to mix in with regular Python projects. You can specify types and other things to help the compiler but, in my experience, even an unchanged Python program becomes an order of magnitude faster when compiled with Cython.

I am with Russell, in hoping that the scientific community sticks with Python for the foreseeable future... Some corners of academia are increasingly enamored with R, which I think is a horrible language (despite its excellent coverage of all things statistical, which explains its popularity).

Telmo
--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.

ronaldheld

unread,
May 6, 2020, 3:28:13 PM5/6/20
to Everything List
Sounds as if I can pass up using it for now.

Philip Thrift

unread,
May 6, 2020, 5:08:51 PM5/6/20
to Everything List


Python can also be used for the so-called higher-level programming things


- stuff like Lisp is known for - which makes it suitable for teaching computer science courses.

It's future is secure. So teach your grandchildren Python.

@philipthrift

To unsubscribe from this group and stop receiving emails from it, send an email to everyth...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages