Neteler: +category

2013-09-09T19:39:28Z

+category

← Older revision		Revision as of 19:39, 9 September 2013
Line 1:		Line 1:
			{{Python}}
	=Why are useful=		=Why Closures are useful=
	A closure is a way to define a variable statically into a function. They are good to:		A closure is a way to define a variable statically into a function. They are good to:
	* Replace hard coded constants;		* Replace hard coded constants;
Line 28:		Line 28:
	<source lang="python">		<source lang="python">
	from grass.pygrass.modules import Module		from grass.pygrass.modules import Module


	def get_module(cmd):		def get_module(cmd):
Line 54:		Line 53:

	Note: This example doesn't work with multiprocessing.		Note: This example doesn't work with multiprocessing.


	=References=		=References=

⚠️Pietro: Add a new page

2013-09-09T19:32:52Z

Add a new page

New page

=Why are useful=
A closure is a way to define a variable statically into a function. They are good to:
* Replace hard coded constants;
* Eliminate globals.

=Some examples=

Define a function that return another function:

<source lang="python">
def print_something(msg):
def _print(var):
print msg % var
return _print
</source>

Now we can use to instantiate the function and define the string formatting statically and then call the returned function.

<source lang="python">
>>> p = print_something('Hi %s!')
>>> p('Pietro')
Hi Pietro!
</source>

Now a more interesting example for the grass users:

<source lang="python">
from grass.pygrass.modules import Module

def get_module(cmd):
mod = Module(cmd)

def do(**kwargs):
mod(**kwargs)
return do
</source>

If we test the speed of a GRASS module, using ipython I got:

<source lang="python">
>>> %timeit Module('g.region', flags='p')
10 loops, best of 3: 30.8 ms per loop
</source>

Instead, using the python closure, I got:

<source lang="python">
>>> greg = get_module('g.region')
>>> %timeit greg(flags='p')
100 loops, best of 3: 12.8 ms per loop
</source>

Note: This example doesn't work with multiprocessing.

=References=
* [http://ynniv.com/blog/2007/08/closures-in-python.html Closures in Python]
* [http://www.shutupandship.com/2012/01/python-closures-explained.html Python closures explained]
* [http://effbot.org/zone/closure.htm Closure]

Python/Closures - Revision history

Neteler: +category

⚠️Pietro: Add a new page