The Bland-Altman plot can be done looking at the ratio A/B as an alternative to the difference A-B.
We're plotting a=mean(X,Y) vs b=Y-X.
In https://rdrr.io/cran/mcr/man/MCResult.plotDifference.html they also allow
a=rank(X)
a=sqrt(XY) [geometric mean],
and
b=(Y-X)/X
b= Y/X
b=(Y-X)/(0.5(X+Y))
Not sure if ALL of these are interesting, but I think Y/X could be interesting in some cases when data is heteroscedastic (error depends on X and/or Y).
The Bland-Altman plot can be done looking at the ratio A/B as an alternative to the difference A-B.
We're plotting a=mean(X,Y) vs b=Y-X.
In https://rdrr.io/cran/mcr/man/MCResult.plotDifference.html they also allow
a=rank(X)
a=sqrt(XY) [geometric mean],
and
b=(Y-X)/X
b= Y/X
b=(Y-X)/(0.5(X+Y))
Not sure if ALL of these are interesting, but I think Y/X could be interesting in some cases when data is heteroscedastic (error depends on X and/or Y).