[GRASS-dev] TGRASS: best practice management of aggregated data

Hi,

while processing MODIS LST time series, I came across some questions
concerning best practice:

Say, I'm aggregating from daily temperature data a set of weekly
min/max/avg maps. So the best pracise is about

- using a reasonable name for the output strds
- using reasonable map names (alias "basename" in t.*)

They should be long-term understandable (with too cryptic names you
will have troubles even a few weeks later) but also parsable (think
"pattern" of g.list).

So, in my case I would call the output strds e.g. "lst_aggreg_modis_weekly".
I would use as "basename" something like "lst_${year}_avg_min_week",
"lst_${year}_avg_avg_week", "lst_${year}_avg_max_week".
(better ideas?)

But: do I put these three sets into a single strds or into three separate ones?

thanks
Markus

Hi,

El sep 27, 2015 9:11 AM, “Markus Neteler” <neteler@osgeo.org> escribió:

Hi,

while processing MODIS LST time series, I came across some questions
concerning best practice:

Say, I’m aggregating from daily temperature data a set of weekly
min/max/avg maps. So the best pracise is about

  • using a reasonable name for the output strds
  • using reasonable map names (alias “basename” in t.*)

They should be long-term understandable (with too cryptic names you
will have troubles even a few weeks later) but also parsable (think
“pattern” of g.list).

So, in my case I would call the output strds e.g. “lst_aggreg_modis_weekly”.
I would use as “basename” something like “lst_${year}avg_min_week",
"lst
${year}avg_avg_week", "lst${year}_avg_max_week”.
(better ideas?)

Have you noticed the -s flag in t.rast.aggregate? It adds date to file names according to granularity of aggregation. I’ve discovered it recently and it is very useful for me.

But: do I put these three sets into a single strds or into three separate ones?

If you use t.rast.aggregate and different methods, you get different strds. IMHO, this is best practice, since you can aggregate them to other granularities without having to select maps by method inside them. Does that make sense?

HTH
Vero


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

On Sun, Sep 27, 2015 at 2:35 PM, Veronica Andreo <veroandreo@gmail.com> wrote:
...

Have you noticed the -s flag in t.rast.aggregate? It adds date to file names
according to granularity of aggregation. I've discovered it recently and it
is very useful for me.

Oh, I didn't. Can you please post an example for the resulting map name(s)?
Then I add that to the manual.

(each run takes me half a day, so I would like to know what -s
produces before using it right away :-).

But: do I put these three sets into a single strds or into three separate
ones?

If you use t.rast.aggregate and different methods, you get different strds.
IMHO, this is best practice, since you can aggregate them to other
granularities without having to select maps by method inside them. Does that
make sense?

Yes! To be added
https://grasswiki.osgeo.org/wiki/Temporal_data_processing#Best_practice_data_organization

thanks
Markus

Hi again

(attachments)

examples_s_flag.txt (3.99 KB)

···

2015-09-27 9:58 GMT-03:00 Markus Neteler <neteler@osgeo.org>:

On Sun, Sep 27, 2015 at 2:35 PM, Veronica Andreo <veroandreo@gmail.com> wrote:

Have you noticed the -s flag in t.rast.aggregate? It adds date to file names
according to granularity of aggregation. I’ve discovered it recently and it
is very useful for me.

Oh, I didn’t. Can you please post an example for the resulting map name(s)?
Then I add that to the manual.

Sure! Please, find it attached as txt and modify as you like :slight_smile:

Cheers,
Vero

On Mon, Sep 28, 2015 at 3:18 AM, Veronica Andreo <veroandreo@gmail.com> wrote:

2015-09-27 9:58 GMT-03:00 Markus Neteler <neteler@osgeo.org>:

Oh, I didn't. Can you please post an example for the resulting map
name(s)?
Then I add that to the manual.

Sure! Please, find it attached as txt and modify as you like :slight_smile:

Thanks a lot, submitted slightly modified here:
https://trac.osgeo.org/grass/changeset/66370

(also to 7.0.svn) and added you to the contributors list.

Best
Markus