Skip to content

Samuel Johnson

Latest
Compare
Choose a tag to compare
@jsmallcombe jsmallcombe released this 09 Mar 20:01
· 32 commits to master since this release
  • Major overhaul of Ultrafit_env and Ultrapeak classes
    • Primary function Ultrapeak::PeakFit() rewritten to be more uniform and fully commented
    • Fitting now performs more careful checks for consistent minimisation and correctly determined error matrix. Fitting steps are repeated with reduced options when problems occur and peak features which fall into problematic ranges are disabled. Additionally the bin integral option is now used for bins width > sigma. Fitting speed for complex multi-peak fits may be noticeably slower, but results will be more consistent an their errors valid. Ultrapeak::PeakFit() is not recommended for use inside intensive code loops.
    • Output of fitting has been streamlined to better highlight any problems users should be aware of before accepting fit results. E.g. parameters at limits or failures of desired fitting options.
    • Ultrapeak main peak maths functions rewritten in a much simplified way which is more transparent to read and cleanly separates the 3 components of the peak (Gaussian, Twin Gaussian and Decay Tail).
    • Added additional cBit options to Ultrapeak class to manage backgrounds and disabling tail component
    • SAVED FITS FROM PREVIOUS RELEASES WILL NOT LOAD CORRECTLY WITH THIS RELEASE
    • Manual controls for TwinGaus parameters added Ultrafit_env
    • Small changes to default parameter limits and "Limit Tail" defaults
    • Added a "no background" option for fitting spectra following interpolative total background subtraction.
  • Added a series of functions to support copy and paste functionality on Linux systems
    • In a jEnv middle clicking on the "Selected Histogram" window will attempt to convert the current clipboard data to a TH1, TGraph, TGraphErrors or TMultiGraph.
    • Currently supports comma and white space delimited plain text as well as data from a LibreOffice spreadsheet copy buffer.
    • A column of text values indicating groups will trigger the TMultiGraph conversion
    • Histogram vs Graph conversion is based on number of rows and the spacing (or lack of) X information
    • Column/Row transposing of input data shouldn't matter
  • Small changes to jEnv for the viewing and drawing of TMultiGraph objects
  • Removed historic excessive use of my name in filenames
  • Added a DOI via zenodo and updated README