> My point is, do people want to actually read values off this sort of plot,
> or just get a qualitative feel for what's happening and where?
Speaking for myself, I'd say both. The original SPECX plot I saw didn't
have them. Tim's original remarks included:
> 3. axis ticks are not drawn except for the first reference plot.
> This prevents things getting too crowded if the plots become small.
> I assume that we could be flexible on this such that ticks are drawn
> if the plot is big enough.
> Again, I just wonder whether people do actually use the ticks marks in
> practice, or whether the borders are sufficient to give your eye something
> to hang on to.
That's partially down to the individual. If you can read those very
long Windows or Ap.J lines without getting a headache then dispense with
tick marks. If you're like me, needs a ruler across inputting tabular
data, then tick marks are a boon. Also the shape and number of plots
affect the decision. If you have say a few wide plots, it becomes
harder for the vision system to connect relative position with
co-ordinates.
So I'm saying we need to be flexible.
Malcolm
|