Regression analysis
Posted: Wed Feb 11, 2009 1:08 am
I'm very pleased with DPlot, but I have a suggestion.
The "Generate polynomial curve fit" function yields an equation that describes a curve fit, which is good, but it would be very useful if it also produced selectable confidence intervals (95%, 90%, etc.) and probability values, so the best fitting polynomial can be selected. As it is, I have to load the original data into a separate statistics program to determine best fit, which is very inconvenient.
The "Generate polynomial curve fit" function yields an equation that describes a curve fit, which is good, but it would be very useful if it also produced selectable confidence intervals (95%, 90%, etc.) and probability values, so the best fitting polynomial can be selected. As it is, I have to load the original data into a separate statistics program to determine best fit, which is very inconvenient.