[GRASS-user] Re: Advice for starting a GRASS project

[Markus Neteler]

Ned,

On Tue, Jan 20, 2009 at 6:25 AM, Ned Horning <nedh at lightlink.com> wrote:
> Markus,
>
> Thanks for the feedback. The advice is very helpful. I've been slowly getting
> used to GRASS but still have a way to go. If I can stick with it I hope to learn
> enough to create a week-long "GRASS for conservation" workshop that we can offer
> at our research station in southeastern Arizona and in our projects in SE Asia.

just keep asking and reading :)

> Here are some clarifying remarks and a few more questions if anyone can help.
>
>> It will work. Everything takes time but I don't expect problems. GRASS 6.2
>
>> has been used for >10GB images, since then more testing and fixing
>> happened. Best is to use the latest 6.4.0RC2 version.

> I'll see if I can get this version set up so it doesn't conflict with the GRASS
> version bundled with QGIS. My linux (Ubuntu) skill are also still quite poor
> but I'm learning (again).

It doesn't conflict. You can have separate GRASS installations
as you like.

>>> 4) Georeference shrub / non-shrub map to reference Landsat images [might
>>> do this with ENVI or ERDAS unless GRASS is a good choice]
>>>
>>
>> Also unclear to me: above you say that you generate the shrub / non-shrub
>> map in GRASS, then it is already geocoded. Or not?
>
> Sorry - I wasn't very clear. The images I'm working with are geocoded but the
> image-to-image registration is poor. I plan to project all of the images into
> the same projection (some are in UTM and some use different datums) and then
> do an image-to-image registration. I wasn't sure if GRASS would be an efficient
> tool to do the image-to-image registration. I should just try it.

There is a graphical geocoding tool available in the user interface.
See
http://grass.osgeo.org/screenshots/imagery.php

>>> 5) Create percent shrub cover map using regression tree algorithm, Landsat
>>> imagery, and shrub location  data from the high resolution shrub / non-shrub map.
>>> I will probably do this using proprietary software unless I can do it easily in
>>> GRASS since a method has been established for another project.
>>>
>>
>> Of course GRASS-R-extention comes to mind for all kinds of statistics.
>
> I'm considering using GRASS for this but would like to know if anyone has used
> GRASS with R as the regression tree engine.
>
> I have a vague recollection of someone
> doing this several years ago but didn't find anything after a quick search.
> I suppose it would require getting training data from GRASS to R to generate
> the regression tree and then port the tree results back to GRASS to apply the
> rules to the dataset. It would be great if an interface already existed to do
> this but if not I can probably figure it out. I'm just not sure how long it
> will take me.

We did a CART some years ago using GRASS and R:
http://www.citeulike.org/user/neteler/article/172938

Best
Markus