floating point / no data source code

[Olga Waupotitsch]

olga at ginko.cecer.army.mil (Olga Waupotitsch) writes:

>We are in a process of updating raster programs, and gis and display
>libraries to process floating point raster maps with additional
>information about data cells.
>Michael Shapiro has documented most of the design and it
>is on my home web page on www.cecer.army.mil
 
>on request of some users I put fp.tar.Z on moon.

NOTE: right now there are 2 files on moon:
fp_main.tar.Z
fp_alpha.tar.Z

>This is a tar file of all the files changed in
>GARSS source tree so far. To unpack it, you have to be in your
>$SRC directory (the one that has display, raster, mapdev,...
>subdirectories)

The above you have to do with fp_main.tar.Z
(i.e. cp fp_main.tar.Z <GRASS_BASE>/src
      cd <GRASS_BASE>/src
      uncompress fp_main.tar.Z
      tar -xvf fp_main.tar
)
For fp_alpa.tar.Z do the same thing, only you have to cd to
your src.alpha directory.
(i.e. cp fp_alpha.tar.Z <GRASS_BASE>/src.alpha
      cd <GRASS_BASE>/src.alpha
      uncompress fp_alpha.tar.Z
      tar -xvf fp_alpha.tar

>After unpacking you need to recompile the whole GRASS
>by running GISGEN.
 
>WARNING: this is grass4.2 code, if you still want to keep
>grass4.1 and apply update packages, you might want to
>make a whole new grass source tree, copy everything over
>and make it grass4.2 as opposed to grass4.1.
 
>It is the best if you have a version of grass4.1 with all 5 update packages
>applied already before you unpack these files on top of it.
 
>All the GRASS commands you are used to will still work
>as they did before except that newly
>modified programs will treat 0's in your old maps as no data cells.
>However all new maps created with modified programs such as
>r.slope.aspect for example, will distinguish between 0's and no-data cells.
>All unmodified prgrams after recompilation will work with new and old
>raster maps the same way they always did.
 
>(See web page for more information)
 
>The source code is in a stage of developement and I am only giving it out
>for people who asre willing to test it. When we change more programs
>here, or fix some problems we might find, I will put an update on ftp site.
 
>Helena Mitasova's group had converted sg.3d plus some of their interpolation
>programs and r.flow to work in the new environment. these changes are not
>included in fp.tar.Z (because they are not part of main GRASS source tree)
>If you need these changes, please email Helena, and she will put them
>on ftp site as well.
 
>Olga