[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: Patterns in an area graph
In principle, you could knit your own area
patterns from point and line elements or by
using area elements as building blocks.
That said, the practical answer is No.
There have been various discussions of
this on Statalist over the years.
In essence StataCorp has long
since cleaved to the line that
area patterns are at root a bad
idea. See the works of Edward Tufte,
In your case you want to contrast
two time series. There would be
seem to be plenty of scope for doing
this via different line patterns
even without colour variations.
> Is there any way I can persuade Stata 8 to draw an area graph with
> hatching rather than a particular colour?
> My coding is below. The key bit is "bfcolor(gray)". The journal we
> would like to submit to publishes in black & white and frowns on grey
> shading because such graphs "do not reproduce well". Fair enough. I'd
> like to replace the area of "gray" with an area of white with black
> dots, or an area of white with black diagonal cross-hatching stripes.
> I can see various commands that control lines, but few that control
> patterns/colours within areas.
> Any ideas? Surely this is not an impossibility?
> Thank you,
> Dominic Muston.
> Here's the code ...
> twoway (area conlow time, base(100) bfcolor(gray) blwidth(thin))
> (area conmid time, base(100) bfcolor(black)
> legend(off) graphregion(margin(small)) xtitle("")
> yscale(range(0 100)) ylabel(0(20)100, grid labsize(large))
> xscale(range(0 15)) xlabel(0(5)15, grid labsize(large))
> name(albgra, replace);
* For searches and help try: