Fitting recommendations

Here are some recommendations and comments on fitting data.

In many cases, to analyze the data recorded during experiments you will need to fit and plot your data.
There are large number of good fitting programs available.

MICROSOFT Excel is marginally acceptable but lacks flexiblity. Fitting is difficult and handling of error bars is weak
Some powerful free programs that work well are "gnuplot" and ROOT. They can be downloaded from the WEB.

"gnuplot" has been used in the past by PHYS480L and PHYS481L students.

("ROOT" is extremely powerful and flexible but may take time to learn. Very popular in high energy and nuclear physics.)

It is extremely important to plot the data with error bars of the correct size,
which reflect the statistical and systematic uncertainties in the measurements.


  • a simple GNUPLOT fit example
  • GNUPLOT home page (much more detailed information and downloads)

  • ROOT page at CERN (includes free downloads, documentation....)
  • Example of setting pointers to directories for root on a MAC

  • Plot data with errorbars in ROOT

  • Fit data with a Gaussian in ROOT

  • Plot two datasets with error bars in ROOT

  • Plotting example with error bars in ROOT
  • Fitting example in ROOT

  • Octave page (includes Wiki and downloads)
  • OCTAVE/MATLAB cribsheet

    Last modified: Sept 23, 2015
    Tom Browder / teb#phys.hawaii.edu