[GRASS-user] best practices for managing locations

Hi all,

I am looking for advice on how to manage my locations. I have Grass installed on my laptop, where I'm the only user. So I use the PERMANENT mapset for each location, which seems to be ok.

However, I've got a growing list of locations that is becoming hard to manage. Each dataset that I download becomes its own location, and then these locations start to get mixed up as I import layers from one location to the other, or create new locations to reproject the original data.

I also deal with a lot of point location data, which gets imported into a lat-long location, then exported out to various other locations. Which introduces a new set of transformations to keep track of.

It's all getting to be a bit confusing. I suspect more experienced users must have developed useful guidelines for managing their own datasets? I hope it's not off-topic to solicit your suggestions here?

Thanks for your time!

Tyler

My method is to make each LOCATION = a coordinate reference system, then each MAPSET holds all layers for some project. So I have a LOCATION called WGS84, and inside several MAPSETs, one for each project. Then another LOCATION called ITM (Israel transverse Mercator) again with several MAPSETs inside. There are cases where I have data for a certain project in two CRS’s so I’ll have two MAPSET’s with identical names in both LOCATIONs.
Works for me…

···

On 27-Apr-14 7:11 PM, Tyler Smith wrote:

Hi all,

I am looking for advice on how to manage my locations. I have Grass installed on my laptop, where I’m the only user. So I use the PERMANENT mapset for each location, which seems to be ok.

However, I’ve got a growing list of locations that is becoming hard to manage. Each dataset that I download becomes its own location, and then these locations start to get mixed up as I import layers from one location to the other, or create new locations to reproject the original data.

I also deal with a lot of point location data, which gets imported into a lat-long location, then exported out to various other locations. Which introduces a new set of transformations to keep track of.

It’s all getting to be a bit confusing. I suspect more experienced users must have developed useful guidelines for managing their own datasets? I hope it’s not off-topic to solicit your suggestions here?

Thanks for your time!

Tyler


grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

This mail was received via Mail-SeCure System.

On Sun, Apr 27, 2014 at 7:41 PM, Micha Silver <micha@arava.co.il> wrote:

My method is to make each LOCATION = a coordinate reference system, then
each MAPSET holds all layers for some project.

Yes, in my group we have a similar concept. The /grassdata/ directory
is on a shared network drive (NFS in our case).

In our projection based LOCATIONs, the MAPSET names are usually
- project names, or
- user surnames (= mapset owner, so they have their personal playground), or
- dataset name in case of larger datasets (MODIS or ECAD time series
with > 10,000 maps)

Works pretty well.

Added to
http://grasswiki.osgeo.org/wiki/Location_and_Mapsets#Best_practice_examples

...

There are cases where I have data for a certain project in two CRS's
so I'll have two MAPSET's with identical names in both LOCATIONs.
Works for me...

Similar here.

Best
Markus

On 27/04/14 04:13 PM, Markus Neteler wrote:

On Sun, Apr 27, 2014 at 7:41 PM, Micha Silver <micha@arava.co.il> wrote:

My method is to make each LOCATION = a coordinate reference system, then
each MAPSET holds all layers for some project.

Yes, in my group we have a similar concept. The /grassdata/ directory
is on a shared network drive (NFS in our case).

Thanks both for your suggestions. That gives me a start in rationalizing my sprawling location tree!

Tyler