Path: csiph.com!newsfeed.hal-mli.net!feeder3.hal-mli.net!newsfeed.hal-mli.net!feeder1.hal-mli.net!newsfeed.xs4all.nl!newsfeed4.news.xs4all.nl!xs4all!post.news.xs4all.nl!not-for-mail
Date: Sat, 15 Feb 2014 13:04:43 -0500 (EST)
From: Ray Cote <rgacote@appropriatesolutions.com>
To: Rita <rmorgan466@gmail.com>
In-Reply-To: <CAOF-KfiD=UgMg7GHiiPiT47zETJmrkicX4YTrxfh4=T2FSq64Q@mail.gmail.com>
Subject: Re: inheriting a large python code base
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Cc: python-list@python.org
Precedence: list
Reply-To: Ray Cote <rgacote@AppropriateSolutions.com>
Newsgroups: comp.lang.python
Message-ID: <mailman.7014.1392488049.18130.python-list@python.org>
Lines: 98
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: csiph.com comp.lang.python:66467

Hi Rita:

Some personal thoughts on how I've approached similar situations:


Step 0: Put all that code under version control immediately.=20
Being able to track and back-out changes is extremely useful -- even in a s=
mall, single-coder app.=20
With something this big, you'll find "things suddenly stopped working" as y=
ou make changes, and
being able to back-track your changes is invaluatable.=20


Step 2: Ask, "Why do you want to optimize the code (run faster)?"
  - what is slow?
    - starting?
    - loading data?
    - solving specific problem?
    - running specific functionality?
    - specific calculations?
    - network access?
    - database access?
    - small data sets?
    - large data sets?
  - who says it is slow?=20
    and what is their definition of sufficiently quick?


3: You state you do not understand the internal logic.=20
I've also inherited applications where the internal logic was unclear.=20
Writing unit and functional tests (if they do not exist) will help to ensur=
e you do not change the behavior of the application.=20

You may find that understanding the internal logic is the best way to speed=
-up the application.=20
You'll likely find that the biggest speed improvements come from optimizing=
 the logic -- not the Python.


4: This is a personal prefernce when inheriting large code chunks -- and I'=
m sure some will counter it is not optimum.=20
Go through the code and run pep8 and pyflakes (or flake8) on all the code a=
nd address the issues that come up.=20
Aside from bringing the code into styling conformance, I find that this hel=
ps me to familiarize myself with the code.=20
As you're making style changes, review individual functions and add documen=
tation if it is missing.=20
If you're not clear what a function does, mark it as such so you know to lo=
ok at it carefully in the future.=20


5: Now you have a clean set of code -- most of which you've at least visual=
ly scanned -- and you're=20
ready to create some timing tests and dig into specifics.

--Ray


----- Original Message -----
From: "Rita" <rmorgan466@gmail.com>
To: python-list@python.org
Sent: Saturday, February 15, 2014 12:10:33 PM
Subject: inheriting a large python code base



hi all,=20


i just inherited a large python code base and I would like to optimize the =
code (run faster). The application is a scientific application so I really =
don't understand the internal logic.=20


Here is what I have done so far, profile (cProfile) and was able to see whe=
re the code was taking the longest time.=20



I was wondering if anyone else experienced something like this and how they=
 overcame a problem such a this=E2=80=A6what techniques have you used to "o=
ptimize" the code.=20






--=20
--- Get your facts first, then you can distort them as you please. --=20
--=20
https://mail.python.org/mailman/listinfo/python-list

--=20
Ray Cote, President
Appropriate Solutions, Inc.
We Build Software
603.924.6079