Pushed to zunzuncode/zunzunsite
c3cb67d Removed unused code for weighted fit statistics, accounted for futureWarning for array comparison to None
Welcome to zunzunsite, a Django site in Python 2 for curve fitting 2D and 3D data that can output source code in several computing languages and run a genetic algorithm for initial parameter estimation. Includes orthogonal distance and relative error regressions. Generates PDF files and surface animations. Based on code from zunzun.com. You will need to install scipy, matplotlib, django, beautifulsoup, imagemagick, gifsicle, reportlab to run this software. On Debian and Ubuntu, you can run: apt-get install python-scipy python-matplotlib apt-get install python-django python-beautifulsoup apt-get install python-reportlab python-reportlab-accel apt-get install imagemagick gifsicle then install the pyeq2 fitting code with these commands: apt-get install python-pip pip install pyeq2 You can now cd to the project's top-level directory and run the django development server with the command: python manage.py runserver and open the url http://127.0.0.1:8000/ in a browser. Cool! NOTES: the code uses Unix-style process forking, and this is not available on the Windows operating system. My tests show that while both mod_wsgi and gunicorn work fine for Django production servers, the uwsgi process model would not allow os.fork() calls to work as required for this software.