Inset within subplot : -D*anchor* is lost

[gd-a]

Inset not taking into account the subplot environment:

gmt begin test png
	gmt subplot begin 3x1 -Fs5c  -R0/1/0/1
	gmt subplot set
		gmt basemap -Bafg
	gmt subplot set
		gmt basemap -Bafg
	gmt subplot set
		gmt basemap -Bafg
		gmt inset begin -DJTR -Rg -JG0/-90/1c -F+gred
			gmt basemap -Bafg
		gmt inset end
		gmt inset begin -DJTR+o1c -Rg -JG0/-90/1c -F+gblue
			gmt basemap -Bafg
		gmt inset end
	gmt subplot end
gmt end show

[PaulWessel]

On the road so cannot check but guessing that 2 insets in a panel is one too many?

[gd-a]

I'd say it's a deeper problem (using -Xa under the hood or something)

[PaulWessel]

I was able to figure out what is going on:

1. No projection is defined for subplot. While we know you mean -JX5c since you specified a 5x5 cm panel and gave a -R, we do set under the hood somewhere that scaling. But when the inset command starts it needs to save the projection used for the panel in which the inset will be placed. At that point it decides no -J is set and hence defaults to the GMT 15x15c linear projection. So when it applies that we are suddenly 15 cm out and up.

I will see if I can figure out exactly where this happens and if I can intervene to set -JX5c there instead, but I would have to create that -J string so easier if it already existed via subplot.