0

I'm trying to use Scipy's scipy.optimise.curve_fit to calculate the parameters for a non-linear least squares fit of my data to a function. I have 2-3 response values for each value of my explanatory variable. Each of these individual response values is assumed to be errorless.

I've done a fair bit of forum searching online but I still need help understanding what I should be specifying for the absolute_sigma and sigma arguments and what effect these have on the output variance-covariance matrix, as I need the matrix for calculating confidence intervals for my fitted parameters and confidence bands for my function via the Delta method.

Any advice would be very much appreciated!

damtheduck
  • 1
  • How is it possible that multiple different dependent values at a single value of an independent variable are errorless? Surely only one of them represents the central tendency – Reinderien Mar 28 '23 at 12:05
  • @Reinderien these are readouts from an experiment - each response value is the readout from an experimental unit (i.e. multiple experimental units have been tested at each level of the explanatory variable) – damtheduck Mar 28 '23 at 16:13

0 Answers0