Hello, You can do this pretty easily with the POSITION keyword for function graphics and direct graphics. For direct graphics you also need to specify the NOERASE keyword to keep the first graphic there. Here is a small example of how to do that for both: plot, findgen(10) plot, findgen(10), position = [.3,.1,.4,.2], /noerase p = plot(findgen(10)) p2 = plot(findgen(10), /current, position = [.5,.3,.6,.4])

Hi, Thanks for the suggestion. However, my main problem is that the base figure is a map. Then the map and the "insert" figures are not with the same scale. The idea is to be able to change the map to any x- and y-range and put the insert figures to their respective location. I am trying to include a figure with this request to explain what I want but the attachment window does not allow me to insert anything. Thanks, A.

Hi Andry, One method for new function graphics is to use the ConvertCoord method which can be used to convert from your data coordinates to device coordinates (pixel x and y). From here, you will have an x-y pair for one of the corners of your custom graphic in the overall graphics window. This is useful because you can set the DEVICE keyword in your next call to PLOT and the value for the POSITION keyword will be given in pixels which you will have already. The only thing you might need to play with is the pixel size of each plot. If that doesn't make sense I can come up with a small example for you. I hope that helps!