HippoDraw canvas window
HippoDraw provides a highly interactive data analysis environment. It is written in C++ with the Qt library from Trolltech. It should run anywhere Qt runs which is just about everywhere. This release has been tested on Red Hat Enterprise Linux 3 and 4, Fedora Core 5, Microsoft Windows XP, and Mac OS X 10.4 both native and with X Windows.
HippoDraw can be used as a stand-a-lone application or a Python extension module. In the latter case, Python is effectively the scripting language for HippoDraw.
Here are some User comments on HippoDraw.
This page gives an overview of HippoDraw. Go to the Quick links to find a short cuts to tutorials, user guides, examples, etc.
HippoDraw has highly interactive controls on the data plots via a dialog window called the Inspector. For example, one can change the width of the bins of a histogram via a slider and see the histogram plot re-display itself many times a second. See Inspector Users Guide and Canvas Window Users Guide for a description of the available controls.
A data source for HippoDraw is typically a NTuple, but HippoDraw supports a number of different kinds described in the data source page. When used as a Python extension modules, Python numerical arrays are also supported. A number of visual plot types can be made with this data. They are typically placed on a canvas as shown Canvas Window Users Guide.
Doing easy data analysis tasks should be easy and they are done with HippoDraw's Inspector. Doing more difficult tasks should be possible. Here, one would expect to find a command line and/or scripting interface to HippoDraw. But HippoDraw does not have either. Instead one of HippoDraw's Python extension modules is used. The Python extension module tutorial allows user manipulation of the data representations as if they were Python objects. This can be done either interactively, or via scripts written in Python. Changes made via Python are reflected in the Inspector and vice versa.
One can use HippoDraw's C++ library for Building custom applications in either C++ or Python. See Examples of custom applications with HippoDraw. The library consists of over 100 classes with about 50K lines of source code. The library is in two parts the core and the GUI part. With debugging symbols, they are about 10 MB and 5 MB respectively with all the optional support enabled. Without debugging symbols, they are under 1.5MB and 1MB. Both can compiled and linked in just over 6 minutes with GCC 4.1.0 on a 2 Ghz Pentium M laptop. A descriptive of the library begins is here. A mirror of the CVS archive is browseable on the Web.
For more information