playground.fitting_statistics_demo

Getting Started

mcphysics.playground.fitting_statistics_demo provides a graphical user interface that simulates the acquisition and fitting of data having Gaussian-distributed error bars. With this, you can specify the source data's underlying "reality" function, the magnitude of the measurement noise, the number of points, a fit function, and the error bars to assume during the fit. It has a loop feature to run many "identical" acquisition / fit cycles, collecting the results and sending them to a histogram / correlation map for inspection. Do the fit error bars correspond to the actual distribution of fit values? How much does the reduced chi-squared value fluctuate?

To create an instance,

import mcphysics
sc = mcphysics.playground.fitting_statistics_demo()

More to come...