Turn-by-Turn Data Analyzer

The application is a graphical user interface for a set of independent component analysis algorithms extracting a useful information from turn-by-turn BPM measurements.

Download: application archive (3.3 MB)
(type "chmod 744 run_in_bg tbt.tcl" after extracting archive, then "./run_in_bg" or "./tbt.tcl")

To store temporary files in a directory other then "(program_directory)/tmp/" type "export TBT_TMPDIR = (desired_tmp_directory)"

Required software

Tcl/Tk interpreter together with BWidget and Img packages may be installed as an Active Tcl distribution.

Algorithms

All signal separation algorithms are implemented as matlab/octave scripts. EVD, AMUSE, JADE algorithms are taken from the ICALAB package. FastICA comes from the Helsinki University of Technology site. EVD and AMUSE algorithms are good at betatron motion separation with closely spaced or overlapping frequency spectra, though they sometimes fail to separate slow changing signals like synchrotron oscillations and ground motion. JADE and FastICA are better for slow changing signals but they cannot separate betatron modes with closely spaced frequencies.

Input data

All input data files are in SDDS format. There are three input files required to run the data analysis: BPM data file, BPM positions file and model (reference) optics file.

Data converters for Tevatron

Examples (13 MB) of usage for all the converters are available.


© 2007 Alexey Petrenko