Re: [R-sig-Geo] Tiled processing...

2008-10-30 Thread Jonathan Greenberg
Ok, I'm nearly there using only RGDAL and R-base commands, if I can get a bit of feedback I might have a decent tiled (row-by-row) processing structure worked up. Couple of things -- first, I was mistakenly thinking all raster formats can even really support line-by-line writing, which is not

Re: [R-sig-Geo] Tiled processing...

2008-10-30 Thread Alexander Brenning
Hi, maybe the RSAGA package has the solution to your problem; there are actually two ways of applying functions to grids in RSAGA, (1) by row-by-row processing (special form of tiles), or (2) using the SAGA binaries. 1) local.function local.function and focal.function are very flexible tool

Re: [R-sig-Geo] Tiled processing...

2008-10-30 Thread Ashton Shortridge
Hi Jonathan, I use GDAL to read subsets of raster and image files, though usually I employ the python bindings. I don't actually have answers, but maybe ideas about how to get there... > 1) The code doesn't actually seem to work -- I get a tiny, unreadable > TIFF out of the back of this algorit

[R-sig-Geo] Tiled processing...

2008-10-29 Thread Jonathan Greenberg
I've recently got back into using R to perform spatial analyses, and I'm trying to figure out how to perform "true" tiled processing, e.g. controlled reading of subsets of an input file, performing a function on this subset, and writing the output, subset by subset, to an output file and, final