I would like to fit an experimental curve using 3 other experimental curves . We don't know the algebric function of the 3 curves, we have only these data, so an x column and 3 y columns. The best fitting functions obtained must be a linear combinations of the three base lineshapes.
Since Origin 9.1, a Rank Model tool has been implemented to allow you fit your dataset with multiple functions and rank the fitting results to obtain the best model.
These tutorials are very interesting. My 3 curves are not described by a known mathematical function but only by a x (the same for the 3) and y (different for the 3) columns.
What exactly do you want to fit? Fit all three Ys with a same fitting function or concatenate three Ys into 1 curve and fit it?
As I mentioned before, if you are not sure what function is suitable for your data, you can run fitting for a list of functions and choose the best fitted model.