Weighting - for regression models

Weighting – for regression models

Weighted regression – this may be used to weight the points in a modelfit.

ch.Set 'Weighting' 1 1 1 1 0.1 1

This would be appropriate if the data was aggregated, but with differing numbers of measurements contributing to each point. Also it may be used to exclude selected points from a model by simply weighting them as zero.

For points with errorbars, you might consider a weight varying as the sqrt of the error to have more certain points counting higher in the model.