example:
http://bambi.otago.ac.nz/hamish/grass/screenshots/inlets_03_SurfSal_icw_big.pngsample data points are given with black diamonds. interpolation
must travel around islands and along channels.
ooh, that looks good. Very useful for interpolating data with faultlines,
I guess. However, quite the overkill for my little problem
by AA, do you mean like a 0 floor and 255 ceiling and depending
on how much vector line in the cell the output raster value
is somewhere in between (with length of cell diagonal == *1.0 ),
or is it still binary null || line_value but in a more smeary
way than normal?
I think suggestion #1 should be implemented in v.to.rast anyway,
as it would be a functional improvement for many use cases where
attribute values from vectors are to be transferred to a raster
with DCELL/FCELL type cells.
Your second suggestion would be an alternative, additional
method with its own uses. Easily implemented, I imagine,
by letting the user control the threshold (% of cell area
crossed by line) above which a cell should be set to "1"
(or whatever). In my scenario, a pixel should be placed,
whenever the % crossed area > 0.0.
The value to place could be the line value or an average of
neighbouring cells. Both may have applications.
Best,
Ben
------
Files attached to this email may be in ISO 26300 format (OASIS Open Document Format). If you have difficulty opening them, please visit http://iso26300.info for more information.