1) The dependent variable are really percentages, my mistake.
2) The independent variable values should be x+273.15 (i.e. the thermodynamic temperature), my mistake again. That explains the parameters from Excel solver.
3) Statistics/Fit had a problem with a syntax:
but anyway I got really similar results from both Statistics/Fit and DirectSearch/DataFit methods.
Now I want to try to fit my data set with a model in logarithmic form:
In this case the the data set values are:
and the k, E parameters should be very similar to the previous fit. However Statistics/Fit and DirectSearch/DataFit claim that "complex values encountered". How can be the fit of logarithmic model obtained?