The best place to see the goals of this project are by looking at the releases in CodePlex and the Issue Tracker as well. Below are listed the basic roadmap – which is always subject to change based on what I need to do in order to get plots published!

Version 0.0

Proof of principle – hacked and rather hard coded. Completed early January 2011. This code was never made public and wasn’t even used for version 0.1 except as an example of how to (or how not to) do something.

Version 0.1

Take version 0.0, and make it more robust, unit tested, and general.

  • Target TTree’s that contained user build objects (like a BJet that inherits from TLorentzVector).
  • Counting of # of objects
  • Make plots
  • Run on local ROOT files (TFile’s).

This was completed around the end of February 2011.

Version 0.2

The main thing this version is looking for is being able to run on different TTree styles

  • Properly deal with a TTree that contains a list of related vector<>’s by falsely creating an object with leaves – basically taking a group of arrays and masking them as an array of objects each containing the items.
  • Make it possible to have an one variable as an index (or array of indices) that point into another object.
  • Drag & Drop the .root file onto a parser algorithm that would then make the XML files that drive the above translations
  • Allow renaming in .NET of the various leaves
  • The start of a helper library to reduce some of the boiler plate ROOT code.

This work was completed around the end of April, 2011.

Version 0.3

This version moves towards a more production version

  • Ability to combine several requests into a single job so that 100 plots can be made in a single go, rather than one result per go (as above).

Version 0.4

Major upgrade focused on analysis speed, and capability.

  • LINQ operators to make common algorithms in HEP easy to code up (matching tracks to jets, or finding two closest together jets, etc.)
  • Analysis Speed Improvements
  • Ease of use improvements (in the Helper’s library).

Version 0.5

Speed improvements

  • Add PROOF or parallel streams support of some sort – whatever gets us more speed
  • Move to a NuGet distribution model
  • Further helper library improvements

Version v.Next

 

I would imagine (hope!) that once 0.3 is produced that everything is ready for 1.0. Among other things, distributing this by nuget and similar will be things that I’d like to see.

Last edited Aug 13, 2011 at 7:39 PM by gwatts, version 3

Comments

No comments yet.