Graphite (v0.69)
This directory contains an release of Graphite, an open-source,
Python graphing/plotting package.
It contains a library and two command-line applications that will
let you make line plots and scatter plots.
It's still in
development, which means three things:
1. Many planned features are not yet implemented.
2. Many implemented features are not yet as easy to use as we'd like.
3. Some implemented features have bugs.
4. Documentation is mostly missing.
However, it also means:
5. It runs and can make cool graphs, and you can have fun playing with it now.
The point of releasing it at this stage is that, as an open source project, we'd like to get as many people hacking on it as possible. We feel it's far enough along to give you a credible promise of greatness, a promise we can easily achieve with your in put and assistance.
But, documentation is sparse, and test cases are too few. Using it as a library (as opposed to the supplied script) will require a bit of rummaging around. We are happy to help.
Features
For an overview of all the features which are planned:
http://Graphite.sourceforge.net/featureoverview.html. (old)
For a detailed list of features which are included in this release:
http://Graphite.sourceforge.net/features.html. (old)
For release notes see:
http://Graphite.sourceforge.net/releasenotes.html. (old)
INSTALLATION
Graphite does its cross-platform, cross-media output through the magic of PIDDLE. If you don't have PIDDLE, you can't run Graphite. Stop now, and go get piddle from http://piddle.sourceforge.net .
Then, you need either the Numeric (Numpy) or numarray modules for python. Numeric is heavily tested; numarray less so.
If you can find the TableIO package for python, it will speed-up operation on large data files, though graphite defaults to using a pure-python read routine and will operate identically without it.
Once you have these installed and working, then you're ready for Graphite. We don't support graphite for python before 2.2, and it's only tested on Python 2.4.
You should be able to execute
python setup.py install
from this directory, and have everything put in its proper place.
If you need superuser access to install, it may make more sense to do
python setup.py build
su
Password: XXXXXX
/usr/local/bin/python setup.py install
^D
so that you don't have problems with compiler path names, if your environment isn't complete for root.
Next, try it out.
This installs a program (g_multiplot) in your path, and you can use it to plot multi-column datafiles. Run it without arguments to get a page of help and usage information.
You can also use graphite as a library. Move to the tests subdirectory, and execute "testa.py". An Adobe Acrobat (PDF) file should appear. View this with Acrobat Reader (or ghostview or whatever), and admire the pretty graph.
You can try the other test programs too. And if you edit testa.py, you'll find that it actually contains a number of different test functions, each producing a different sort of graph. Go to the bottom of this file to select which test is actually run. You can also change which PIDDLE backend is used to render the graph at the top of the file. Experiment. Let us know how it goes.
Getting Started
To get started check out the tutorials on the web page.
http://Graphite.sourceforge.net
Also, read the Graphite white paper.
CONTACT INFO
Graphite was initially conceived by Joe and Michelle Strout, with extensive helpful input from David Ascher and other members of the Python plotting community. We expect Graphite to continue developing rapidly, so you'll want to stay on top of things by frequenting the Graphite web site:
http://Graphite.sourceforge.net/
and by subscribing to the mailing list:
http://lists.sourceforge.net/mailman/listinfo/graphite-list
Good luck, have fun, and stay in touch!