[GRASS-dev] Question on r.grow 'metrics' parameters

GRASS GIS GRASS GIS developer
1

I was wondering if anyone could describe, or point me to a link of, the various r.grow metrics parameter definitions: euclidian, maximum, and manhattan.

Also, are the comments regarding how r.grow decides how to grow between two equally eligible cells here (http://www.nabble.com/-GRASSLIST%3A7611--r.grow-to8629593.html#a8629593) still valid? I would like to add this information to the docs, and AFAICS, main.c hasn't changed much since Glynn's last edit 21505 back in late November, 2006.

Thanks,

~ Eric.

2

I found some of the metrics explained in wikipedia under Taxicab geometry

Helena

On May 16, 2008, at 12:29 PM, Patton, Eric wrote:

I was wondering if anyone could describe, or point me to a link of, the various r.grow metrics parameter definitions: euclidian, maximum, and manhattan.

Also, are the comments regarding how r.grow decides how to grow between two equally eligible cells here (http://www.nabble.com/-GRASSLIST%3A7611--r.grow-to8629593.html#a8629593) still valid? I would like to add this information to the docs, and AFAICS, main.c hasn't changed much since Glynn's last edit 21505 back in late November, 2006.

Thanks,

~ Eric.

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

3

Helena:

I found some of the metrics explained in wikipedia under
Taxicab geometry

see http://article.gmane.org/gmane.comp.gis.grass.devel/2330/

with manhattan AFAIU you can go crosstown or up/downtown but not diagonally through sky-scrapers.

(ignoring the mess of cobblestones below 14th st + broadway :slight_smile:

Eric:

> I was wondering if anyone could describe, or point me
to a link of,
> the various r.grow metrics parameter definitions:
euclidian,
> maximum, and manhattan.
>
> Also, are the comments regarding how r.grow decides how to grow
> between two equally eligible cells here
(http://www.nabble.com/-GRASSLIST%3A7611--r.grow-to8629593.html#a8629593)
still valid? I
> would like to add this information to the docs, and AFAICS, main.c
> hasn't changed much since Glynn's last edit
> 21505 back in late November, 2006.

don't know. but beware there was r.grow and r.grow2 which worked differently.

Hamish

4

Patton, Eric wrote:

I was wondering if anyone could describe, or point me to a link of, the
various r.grow metrics parameter definitions: euclidian, maximum, and
manhattan.

Euclidean (aka "Pythagoras' theorem"):

  d(dx,dy) = sqrt(dx^2 + dy^2)

  http://en.wikipedia.org/wiki/Euclidean_metric

Isolines of distance from a point are circles

Manhattan:

  d(dx,dy) = abs(dx) + abs(dy)

  http://en.wikipedia.org/wiki/Manhattan_metric

Isolines of distance from a point are rhombi (diamonds).

[See Hamish' reply for the reason behind the "Manhattan" name.]

Maximum:

  d(dx,dy) = max(abs(dx),abs(dy))

Isolines of distance from a point are squares.

Also, are the comments regarding how r.grow decides how to grow between
two equally eligible cells here
(http://www.nabble.com/-GRASSLIST%3A7611--r.grow-to8629593.html#a8629593)
still valid?

Yes.

--
Glynn Clements <glynn@gclements.plus.com>