As python 3.X becomming more and more popular,Can we developers take developing biopython that support python 3.X into consideration? I am a newer to biopython and find that biopython doesn't support python 3.X. It's really frustrated. Thank you !
•
written
by
Young •
There's a good chance NumPy will support Python 3 in a stable release sometime this year -- it's a stated goal of theirs, but getting there has been difficult. Once NumPy makes the switch, Biopython and probably a number of other scientific libraries for Python will start making sure 2to3 conversion works properly.
Biopython generally works with the development releases of Python 2.7, so there's a good chance converting to Python 3 with the 2to3 script will be mostly painless.
Python 3 will only become widely used in sciences once numpy (the numerical library) gets ported from the version 2 of python. I would recommend staying with the 2.X line for the next year or two.
In fact, I have started to consider the possibility that the 3.X line will never get the sufficient traction to be popular in sciences.
re numpy, it's pretty well converted:
$ py3
Python 3.1.2 (r312:79147, May 17 2010, 21:50:36)
[GCC 4.4.3] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
>>> numpy.test()
Running unit tests for numpy
NumPy version 2.0.0.dev8417
[... lots of output ... ]
Ran 2929 tests in 15.571s
OK (KNOWNFAIL=5, SKIP=4)
[?]
but yeah, there's not much motivation to switch now.
re numpy, it's pretty well converted:
$ py3
Python 3.1.2 (r312:79147, May 17 2010, 21:50:36) [GCC 4.4.3] on linux2 Type "help", "copyright", "credits" or "license" for more information.
>>> import numpy
>>> numpy.test()
Running unit tests for numpy NumPy version 2.0.0.dev8417
[... lots of output ... ]
Ran 2929 tests in 15.571s OK (KNOWNFAIL=5, SKIP=4)
[?]
but yeah, there's not much motivation to switch now.
Python 3 will see lots of speed improvements in the next years, e.g. by adding Unladen Swallow into the main line. So I do hope that the libraries finish their migration eventually.
@brentp: I did not know the porting of numpy was in this advanced stage, that's encouraging
The notion that what is hampering the usage of Python 3 in science is numpy is borderline ludicrous. After all, we, scientists, should be in the forefront of these type of developments, using things that people do not use still. Python 3 is not widely used in science, because Python 3 is still not used everywhere and not every third-party module is already ported to it. Every time I install a new module, I try to install it under Python 3.
numpy matters because how else would one operate on large vectors and matrices without it?
Looks like NumPy and SciPy are pretty usable with Python 3 now: http://article.gmane.org/gmane.comp.python.numeric.general/39033
Although converting to python 3 is painful and tedious for many of the programmers who are familiar with python 2.However, if we are still hesitating we may lose the chance while other people or oragnisition may develop another biology analytical tool using python 3.
I am a newbie to programming and I recently got the O'Reilly "Bioinformatics Programming Using Python" book. This covers Python 3 and now I am in a fix - whether I should learn programming using Python 3 or should I get an older edition of the book which covers Python 2.x.
But from what I have read above I understand that by the development of the modules for Python 3.x is going on pretty well and also the Biopython github has this message posted - "Python 3 support is still incomplete, but the majority of modules are already functional." So I guess by the time I learn Python the numpy module and Biopython are ready for show time.
Log in to answer this question.
Hopefully they'll pep8 the names too.
Note this question was asked back in 2010, and is out of date. After a period of beta-level support for Python 3, Biopython first officially supported Python 3 with Biopython 1.62 (August 2013) which was done using the 2to3 conversion script, and as of Biopython 1.63 (December 2013) supports Python 2 and 3 directly with the same code base.