I need to compare two non-linear models of growth. The first one is calculated with nls function in R and with non-linear estimation function in Statistica - both programs gave identical results and provide both correlation coefficient R and AIC/AICc values.

However, I've got also second, suboptimal model, presented in older article. Both models are curves of the same type [y=a*exp(b*x)], yet have different "a" and "b" parameters. I want to check, how good is the curve fitting made by previous authors (and how much better is the new model), yet I don't know how to check it's R and AIC in relation to defined dataset. Both nls/non-linear estimation functions are made for finding best possible parameters and do not allow to use custom ones.

Then - how can I do it? Which functions should I use?