Welcome to PARK!

PARK is a simple and easy to use framework for the distributed computing. It is implemented in pure python, and runs under Windows, Linux/Unix, and Mac on desktops, shared-memory multi-core computers, and PC clusters. It can be downloaded as the service manager to manage the distributed services, the distributed fitting service server, or as a component in DANSE environment. Enjoy :)

Posted on November 22, 2007 by Admin

Lastest Announcements

The beta 1.0 of PARK is released for the testing.

The following features are added in this release:

  • Unify the interface for the client applications based on script file
  • Provide tools to develop the script client applications
  • Solve the long startup problem to start a new service (30s -> 1s)
  • Add new examples to demonstrate the developing of new model for fitting service
  • A few bugs are fixed in GUI client applications for fitting service
  • Support the optimization for the box-constrained problem
  • Provide several optimization algorithms based on scipy.optimize moduls
  • Provide uncertainty estimation based on Levenberg-Marquardt algorithm for least-square fitting problem
  • Separate the data structure and GUI representation for the fitting service
  • Finish the first draft of the tutorial
  • Provide API document generated by epydoc

Posted on November 22, 2007 by Admin

Lastest News

The beta 1.0.1 of PARK is released for the testing.

Posted on Febuary 15, 2008 by Admin

Financial Support

Sponsored by NSF grant DMR-0520547.