[GRASS-dev] terminology fixes in GRASS

[Markus Neteler]

Helena Mitasova wrote on 12/13/2006 07:04 PM:
> Markus,
>
> although I suggested the change, when I browsed the man pages for the
> context
> how this terminology is used, there are few instances
> where raster file may be justified - for example if you are importing
> an external raster file by r.in.* to create GRASS raster map,
> although increasingly also the raster to be imported would have multiple
> files so then raster map would be OK.

Helena,

exactly these cases I left untouched. I didn't bulk-replace, but verified
each change. There is a low probability that I got 1-2 entries wrong,
please complain in this case.

Example:
grep 'raster file' raster/r.in.*/*
raster/r.in.arc/main.c:         _("Convert an ESRI ARC/INFO ascii raster
file (GRID) "
raster/r.in.arc/main.c: parm.input->description = _("ARC/INFO ascii
raster file (GRID) to be imported");
raster/r.in.arc/README:Module to import ARC/INFO ascii raster files
raster/r.in.ascii/main.c:         _("Ascii raster file to be imported
(or \"-\" to read from stdin)");
raster/r.in.bin/main.c:     _("Import a binary raster file into a GRASS
raster map layer.");
raster/r.in.bin/main.c: parm.input->description = _("Binary raster file
to be imported");
raster/r.in.gdal/main.c:        _("Import GDAL supported raster file
into a binary raster map layer.");
raster/r.in.mat/main.c: * Input a GRASS raster file from a MAT-File
(version 4).
raster/r.in.xyz/main.c:    /* close raster file & write history */


>
> In general, when making this change it is good to
> check whether there may be a reason for having a file rather
> map to distiguish between a raster stored in a single file rather
> than multiple files or there may be some other reason too
> that I might have missed.

I tried to do so.

> Maybe some users could check the updated man pages to see
> whether the change raster map might have caused a problem.

Verification is of course welcome.

>
> For 3d we should probably use 3D raster map to be consistent.
> Just an explanation - in literature grid is used for raster that
> represents continuous field and the value is applied to the point
> and you need interpolation to resample, raster would be image
> or discrete geometry (e.g. land cover, soils) where we have cats
> associated with pixel (area) and you use nearest neighbor to resample.
> Most volume data would be continuous
> 3D field, that is why the grid was used.
> But I think that for manuals and help we are fine with calling everything
> raster and 3d raster as we don't need to make the terminology
> more complicated for users than it already is.
>
Thanks for your comments,
Markus