Quasars / orange-spectroscopy

Other
51 stars 58 forks source link

Improving Resolution Fit #631

Open yjohar opened 2 years ago

yjohar commented 2 years ago

Hello,

Hello, I am relatively new to Orange(and data mining) and wanted to ask for advice. I am fitting a bunch of spectral data to Fano (see picture). The fits are pretty sharp, partly due to the resolution of the instrument used to collect data. My question is if there is a way to improve "the resolution" of the fit to make the fit smoother. My guess would be "adding" points without affecting the data to make it smoother but I am not familiar with this. Please let me know what can be done!

image

stuart-cls commented 2 years ago

Hi @yjohar , thanks for your message.

You are correct, the data spacing of the fit plot is defined by the input data spacing. However, the widget outputs the best fit model parameter results for each curve, which of course do not contain any information about the data spacing.

We don't have a great way to visualize fit results outside of the widget right now. In principle I think you could create a second widget with fixed parameters equal to the output of your fit and then feed finer data spacing data into that (I suggest using the Interpolate widget). In this way you could generate a finer-grained output of the fit and then combine with your original data in some plotting software.

Sorry I don't have a more elegant suggestion at this time.