fast mapcalc

[Jim Hinthorne]

>
>Hi !
>
>As some people might know, I been using Grass for the last 7 years.  
>Unfortunally not so much the last two years as before due to my current 
>job. But right now I have a problem that might leed me into using grass
>more activly again.
>
>I wonder if someone have thought about doing the kind of processing as 
>r.mapcalc could do for you but the result would still be primary memory
>in between different operations.  What I would like to do, is to apply
>operators on a image that I am working on without having to save the result
>on a disk file. I guess xv is working a little this way even if I haven't
>looked into the details of xv. What I also would like to do is to be able to 
>run r.stats on the same image that I still only have in memory not on disk.
>At last, I am NOT interested to se the result on screen.  The result I am 
>interested in, is the statistics from r.stats.
>
>I would be very glad for some comments on this and possible if someone have
>modified grass in this way.  Or maybe there is some other public domain
>software that already is doing this. 
>
>Lars
>
>Lars Schylberg                       Email: lasc at celsiustech.se
>CelsiusTech IT                          or: larss at fmi.kth.se
>S-175 88 Jarfalla                    Tel.   +46 8 580 847 03
>Sweden                               Fax.   +46 8 580 123 20
>
>
>


Most Unix workstations allow you to define a ram-disk.  If you have enough 
RAM you can copy your location to the ram-disk and then run grass normally.

Jim Hinthorne              Voice:  509-963-2826
GIS/Geology                FAX:    509-963-1047
Central Washington Univ.   E-mail: geol1 at cwu.edu
Ellensburg, WA 98926-7418