Very pretty output. The only thing I'm missing in the output is a CI of the LoA, and an option to fit a regression line in the graph in cases where the measurement error differs when the mean is high (or low) (eg. funnel shaped or non-horisontal B-A plots). Otherwise your program has all the basic features MedCalc provides https://www.medcalc.org/manual/blandaltman.php
thanks seraina, this is really helpful. By comparing means so the distribution has to be normal, am I right? Can you tell me the interpretation of limits of agreement?
Seraina Caviezel I don't get any data plots on my graph when executing the command, do i have to edit your script in any way for the graph to display properly?