|
From: Michael L. <M_L...@ho...> - 2005-07-14 03:09:12
|
Hans-Bernhard Broeker <broeker <at> physik.rwth-aachen.de> writes: > > Michael Lasarev wrote: > > -snip- > > What I want is a sharp distinction between the boundaries but nice > > color gradients in each submatrix showing the transition towards +/- > > 1. > > That wish is quite self-contradictive. > I wrote that poorly, you're right. I meant that I wanted a clear distinction between the positive- and negative-valued blocks of the matrix but with a gradient within each positive (negative) block showing how the values transition towards +1 (-1). [They are already close to the boundaries with absolute values > 0.8, mostly.] For instance a blue gradient within the positive-valued block and a red gradient in the negative-valued block, but without a 'zero colored' margin where it transitions from -0.8 to +0.8. corners2color did help, thank you. I also had good results by converting the matrix into three separate columns (x-position, y-position, z-position=correlation) and then splot 'three_cols_of_data' u 1:2:3:3 with points pt 9 ps 0.70 palette --Mike |