The cell (i,j) contains the distance between the i-th and j-th color of the color map, which are displayed on the diagonal.

The cell (i,j) contains the ratio of the distance between the i-th and j-th color of the color map and the difference of their corresponding values. The higher the global speed, the better global discriminative power of the colormap.

The height of each column i shows the ratio of the distance between the i-th and (i+1)-st color of the color map and the difference of their corresponding values. The higher the locaö speed, the higher is the colormap's' local discriminative power in this area.

The cell (i,j) contains the asolute value of the mimum of |c(i)-c(j)|-|c(i)-c(k)| over all k between i and j (i < k < j). As long as it is positive, the middle color is closer to each of the two outer ones than theses are to each other. That means, the colormap has an intuitive order everywhere between i and j. That is why, we only plot the values that are nageative, i.e. where the global, intuitive order is violated.

If there are k between i and j that violates the intuitive order, we plot the color c(k) that does so the most, i.e. the one that produces the minimum diplayed in the visualization above.

The height of each column i shows the absolute value of the minimum of |c(i)-c(i-1)|-|c(i+1)-c(i-1)| and |c(i+1)-c(i)|-|c(i+1)-c(i-1)|. As long as it is positive, the middle color c(i) is closer to each of the two outer ones than theses are to each other. That means, the colormap has a local, intuitive order in this area. The bars strat at zero.