search

ECSHackWeek / impedance.py

A Python package for working with electrochemical impedance data
MIT License
212 stars 102 forks source link

[BUG] (not sure it is a bug) #232

Open flamariond opened 1 year ago

flamariond commented 1 year ago

Describe the bug Fitting seems to be inadequate for a p(R1,C1) circuit with large values of R (eg 1e+9) e small values of C (eg 1e-12)

To Reproduce Steps to reproduce the behavior:

  1. I have a synthetic data for a p(R,C) circuit in the frequency range of 10^5 Hz to 1 Hz
  2. impedance.py does not fit it. It seems to hold constant one of the parameters.

Expected behavior Since it is a synthetic data, fit should be perfect.

Additional context I have been using this excellent package for a couple of days, and found this difficulty trying to fit data obtained from a painted metallic surface on NaCl 3.5% (it is a corrosion study), with these high values of R and small values of C. Congratulations for this excellent software.

mdmurbach commented 1 year ago

Hi @flamariond, thanks! Would you be able to share the code you have to reproduce this?

adrianusler commented 1 month ago

I also had a problem with fitting impedance data when the parameters were supposed to be very small $<10^{-10}$. This was related to a bug in scipy.optimize at the time: https://github.com/scipy/scipy/issues/18793#issue-1781188767 Could this be related to your problem, @flamariond? Which scipy version do you use?