[GRASS-dev] testing results of r.watershed2 against old r.watershed

[Hamish]

Markus Metz wrote:
> Change option names also for r.watershed.ram and
> r.watershed.seg? There are some options in uppercase.

those do not use G_parser() and are not exposed to the user, so not a
priority for standardization.

Hamish:
> > see the man page for an example of making a nicely colored accum map
> > based on standard deviations.
MMetz:
> Why not setting colors for accum in the module?

If you like, but a simple linear color model will not work well:

r.univar -e for absolute value of accumulation map:

Of the non-null cells:
----------------------
n: 2654802
minimum: 1
maximum: 811721
range: 811720
mean: 643.885
mean of absolute values: 643.885
standard deviation: 12230.5
variance: 1.49586e+08
variation coefficient: 1899.49 %
sum: 1709387991
1st quartile: 3
median (even number of cells): 6
3rd quartile: 14
90th percentile: 32


ie the bulk of map has little flow, but rivers near outlets have lots,
so a statistical method like in the man page example is needed to show
the detail, the standard linear color gradients (or even a log one)
won't do.


> > in my tests r.watershed(2) is >80 times faster than r.watershed(1).

> The speed increase is not static, the new version will be
> faster the larger the region (more cells). For somewhat
> larger regions, the new module is >1000 times faster.

ok, can you suggest better wording for the release announcement?

t>1000 may as well be infinite, as before the user hit ^C after 2-3 days
and so it never finished. So this opens the door to analyzing much bigger
(ie modern) datasets; r.terraflow is nice for those, but doesn't provide
the catchment basin output maps.


> > With -m it took just under what I set memory= to. If I set mem=950 it
> > used 911mb RAM. Does it not know it only needs ~166mb instead of full
> > alloc?
> >   
> It should, but I made a mistake in adjusting the number of
> open segments. Please apply the diff attached.

tested and applied, thanks.


> I wonder how many more bugs will surface after more testing...

time will tell, but I think it's pretty good,


cheers,
Hamish