This program implements a version of trace fitting inside the ROOT framework. The previous versions of this fit were done with gnuplot. The results between the two different programs/minimizers has been checked and the results are consistent.
The fitting functions included with this program will be integrated into the pixie_scan code developed by the Nuclear Structure group at the University of Tennessee at Knoxville. They will makeup the core routines to perform high resolution timing for VANDLE.
The program now recognizes the following data structure :
- x,y
- x,y,y_error
- x,y,y_error, ... ,n
Only the first two columns will be read into the program. In all of the current data files that I test with the x column is generally the trace sample number, and the y value is the amplitude of the trace at the sampling point. The program will return a y_error vector in the event that one wants to use it for some of the calculations. In principle, this vector is not used and is ignored.
./traceFit
Currently, some of the features are hard coded. For example, the upper and lower bounds of the fit. I currently find it easier just changing them and recompiling than adding them as run time arguments (hooray IDEs). The same goes true for the fitting function to use. I have not found a trivial way to switch this up in the way that I am currently implementing the fitting. If somebody finds something better, let me know, please!