I have a dataset in which both x and y axis are measurements which have an error term associated.

I am making a graph where the relation between both variables is shown.

When I plot the vertical and horizontal error bars, a clear error area arises and I want to make a nice graph of this. However by plotting the error bars, the graph seems overcrowded and actually what I want to show becomes blurred. The graph has a lot of points (about 100). That's why I'm a big fan of showing error areas rather than error bars. For some points the x-error terms is the largest, while for other points the y-error is the largest.

What I am looking for is a function/script/tool to define the largest errors for every x and y coordinate so I get two curves which I can fill with an error band.

Anything that can work would be great. Preferably I am looking for a solution in R, Sigmaplot or Excel, but any suggestion is good.

Similar questions and discussions