I have a data set which looks like an incompleted gaussian distribution, or bell curve truncated on the right. How can I produce a Trendline best-fit Gaussian so that I can get the mean, standard deviation, and R-squared for the fit ?
I have written a QBASIC program (for DOS) that can generate data points for a given mean and SD, that I can then paste into a spreadsheet alongside my data, and perform a sum of squares of differences, but that is only one fit, not "best-fit".
Surely someone has done this before ?
I have Borland Delphi 2005 if that helps, but from the look of it, its going to take a year to master it.