[GRASSLIST:3600] Re: Geophysical framework: WHITE PAPER

[Hamish]

> > I would like to keep the grid layout information intact, because I
> > deem it very handy.
> 
> This is a good idea. I'd suggest keeping it in a simple text file with
> structured variables much like a GMT or ESRI header file. Thus, each 
> data set (i.e., from a grid sample) would be stored in a text file 
> named <filename>.grd (or something like that) and a corresponding grid
> definition file would be stored in <filename>.gdf.


.grd is used for Surfer ASCII grids already.

No need to be tied to 8.3 filenames anymore though. Make it '.grid' ..
'.html' has overcome this habit for the most part, but not much else.


some general notes:

- If you are planning to add directories to the $MAPSET, please do so
following GRASS 5.7's vector/ example to help keep things clean and
tidy in the future.

e.g. for two grids, foo and bar:
$MAPSET/grid/foo/layout
$MAPSET/grid/foo/data
$MAPSET/grid/bar/layout
$MAPSET/grid/bar/data
etc.


- As already noted, r.composite already exists, a new name will be
needed.  r.grid.composite?


- I'd think it useful to be able to produce some field sheets once the
grid is defined. Take an aerial image, oversaturate it, and use it as
the backdrop on a page with the grid drawn as solid black/red lines
with optional Bingo-id numbers drawn in each grid-cell.

5.7's d.m -> Print button could make something like this fairly easily
if the image + vector file already existed.

Zoom/size based on grid extent & A4/USletter, image saturating by
r.mapcalc, & rendering could all be done in a script, producing a set of
rules which could be loaded by d.m's File->Open.

Good for planning in advance; reducing confusion & recording errors in
the field; return trips; or for a classroom exercise.

I don't know how much you folks have to change your plans once on the
ground though..



Michael:
> A new function that would be handy would be the ability to develop a
> color table using the combined values of multiple datasets that have
> NOT been patched together.

Write a script that uses 'r.info -r' on all the maps to figure out the
overal MAX/MIN, create a rules file from those numbers, then apply to
all maps.

Same thing is useful for making movies of a time series.

alt1. use r.univar instead or 'r.info -r'

alt2. make some temporary maps with r.series and then run r.univar on the
results to get the info you need.




Hamish