[GRASS-user] mfilter (comparison with r.neighboors)

Schmitt Thierry thierry_schmitt at yahoo.com
Tue Mar 20 10:38:38 EDT 2007


Thank you for your comments and suggestions.

Am I wrong to think that r.neighbors and r.mfilter, for a 3x3 window for example, should give me the same output. If not could you help me enlighten my comprehension of those two tools.

Cheers
 
 _____
|______\   Thierry SCHMITT
      ||          Géologie Marine et Côtière
        ||          Marine and Coastal Geology
        ||          www.cidco.ca



----- Original Message ----
From: Glynn Clements <glynn at gclements.plus.com>
To: Jarosław Jasiewicz <jarekj at amu.edu.pl>
Cc: Schmitt Thierry <thierry_schmitt at yahoo.com>; grassuser at grass.itc.it
Sent: Tuesday, March 20, 2007 10:16:53 AM
Subject: Re: [GRASS-user] mfilter


Jaros-Bław Jasiewicz wrote:-A

> > I do wonder if mfilter can perfom filtering on float raster, ie
> > providing results as raster. My recent use of r.mfilter led me to
> > the contrary (i.e output is integer also input was raster). I am
> > wrong somewhere?
> 
> well, not
> The only solution is to multipy orginal map before filtering, and divide 
> it after
> 
> but... it is possible to change the r.mfilter to output fp maps? Is I 
> read source code the results is fp and next is rounded to near integer.  
> Maybe removing this part will be enough (but I'm not sure, I'm not 
> skilled enough in grass programming)

There isn't any fundamental reason why r.mfilter cannot be changed to
create FP maps.

Apart from allocating FCELL/DCELL buffers, certain functions need to
be changed, e.g.:

    G_open_cell_new -> G_open_fp_cell_new
    G_get_map_row -> G_get_d_raster_row

Also, the code would need to check for and handle null values; the old
G_get_map_row() interface silently converts nulls to zeroes.

-- 
Glynn Clements <glynn at gclements.plus.com>




__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 




More information about the grass-user mailing list