I need to fit data where the data points are not equally distributed. However, the fit uses a convolution. From my understanding for the convolution to work, the arrays need to have the same spacing. For that I interpolate the data. But now, I fit the interpolated data, and anywhere where I only have few data points, the interpolation gives more datapoints, hence these regions will have an influence on the fit, that is overweighted.

Specifically, I want to fit time resolved absorption data, that have an exponential distribution of data points. I work with python.

Current idea of solving: With python curve_fit, I can give weights, so to counterbalance the overweighing by interpolation. Not exactly sure, how to do that, but I guess it should work.

I just wonder, since this should not be an uncommon issue, if anyone knows the apprpriate way to solve.