Popular recipes tagged "performance" but not "iops"

socket.sendfile() (Python)

2014-06-13T13:41:25-07:00

Python recipe 578889 by Giampaolo Rodolà (backport, performance, python, sendfile). Revision 2.

This is a backport of socket.sendfile() for Python 2.6 and 2.7. socket.sendfile() will be included in Python 3.5: http://bugs.python.org/issue17552

A Simple Timing Function (Python)

2013-12-01T01:39:44-08:00

Python recipe 578776 by Mike Sweeney (performance, timing). Revision 2.

This function prints out a message with the elapsed time from the previous call. It works with most Python 2.x platforms. The function uses a simple trick to store a persistent variable (clock) without using a global variable.

Generator with lookahead (Python)

2013-09-24T10:44:56-07:00

Python recipe 578671 by Rutger Saalmink (background, generator, generators, lookahead, performance, python).

Python generators are a great way of reducing memory usage due to the lazy (on demand) generation of values. However, when the process-time of generating such a value is relatively high, we can improve performance even more by obtaining the next n values of the generator in a separate thread in the background. Hence, the BackgroundGenerator.

Performance boost with metaclasses (Python)

2013-02-17T10:58:13-08:00

Python recipe 578460 by Martin Schoepf (metaclass, performance).

Consider a highly dynamic system where a lot of objects are created and destroyed. If the objects are complex and have a lot of default values the following recipe may improve the performance. The more default values you have the more you will gain from it.

Fast and elegant switch/case-like dispatch (Python)

2011-09-02T01:49:40-07:00

Python recipe 577864 by Jan Kaliszewski (case, decorator, dispatcher, performance, switch). Revision 4.

My approach to that common issue focuses on efficiency and elegant, declarative style of definition. That's why:

The way switches work is based on unwrapped defaultdict/list lookup.
The way you define switches is based on classes and easy-to-use decorators (note that you can use subclassing in a flexible way -- without any negative impact on efficiency!).
Its use cases focus on a-lot-of-keys situations and it does not cover the fall-through feature (though you can reach its semantics if you really need that -- by calling class methods...).

Compare algorithms for heapq.smallest (Python)

2011-12-25T23:41:23-08:00

Python recipe 577573 by Raymond Hettinger (algorithms, benchmark, heaps, performance). Revision 3.

General purpose technique for counting comparisons in various searching and sorting applications.

Using a getter for a one-time calculation of a JavaScript object attribute (JavaScript)

2010-07-16T18:10:51-07:00

JavaScript recipe 577310 by Trent Mick (getter, performance).

This is a technique for using a JavaScript getter to calculate the value of an attribute just the first time. Rather than caching the value in some private variable and returning it, you just delete the getter and put the calculated value in its place.

Note: I'd read about this technique ages ago, but forgot the details and had to look-up "delete" for removing the getter. :)

quick Python profiling with hotshot (Python)

2009-02-20T22:35:07-08:00

Python recipe 576656 by Trent Mick (hotshot, performance, profile). Revision 2.

This is a quick snippet that I use occasionally to profile some pure-Python code, using hotshot. Basically it is this:

Put this @hotshotit decorator on the function you want to profile.
Run your code through some representative paces. The result will be a <functionname>.prof in the current directory.
Process the .prof file and print the top 20 hotspots with the given "show_stats.py" script.

Props to Todd for slapping this code together.

Hotshot is a little out of favour now, so I should -- or Todd :) -- should really come up with an equivalent that uses cProfile.