Hi,
The only way I know to do this is as follows:
1) Get your images and get the matrices of pixels for them.
2) Overwrite a submatrix of the "lower" image's pixels with the pixel matrix of the "upper" image.
Below is a script illustrating this, and an image (actually a window of 3 images where the "top" image has been "pasted" in different locations) showing the results. You may have to experiment to get your image where you want it, but this approach should eventually do the trick.
names default to here(1);
//make a table
dt = astable(J(50,2,randomuniform()));
//make a graph
gb = dt << Graph Builder(
Size( 500, 500 ),
Show Control Panel( 0 ),
Variables( X( :col1 ), Y( :col2 ) ),
Elements( Points( X, Y, Legend( 5 ) ) ),
);
//save to a picture, get the pixels and get the picture's x and y dimemnsions
pic1 = (report(gb)[framebox(1)] << get picture) << get pixels;
pic1y = nrow(pic1);
pic1x = ncol(pic1);
//now make another graph and do the same
dist = dt << distribution(histograms only, y(:col1));
pic2 = (report(dist)[framebox(1)] << get picture) << get pixels;
pic2y = nrow(pic2);
pic2x = ncol(pic2);
//create a window to place the resulting images
nw = new window("xx",
vlb = vlistbox(
)
);
x = y = 1; //initial "corner" of overwritten pixels... we will increment these and do 3 total pictures
for(i = 1, i<=3, i++,
newPic = pic1;
newPic[ y::(y+pic2y-1), x::(x+pic2x-1), ] = pic2;
vlb << append(picturebox(new image(newPic)));
y += 60;
x+= 150;
);
Cheers,
Brady