Select to view content in your preferred language

# Judging Polygons regarding Compactness Ratio

2631
9
02-22-2013 11:34 AM
New Contributor III
Dear All,

while analysing polygons respective their fitness of size and form to build power plants there I am facing the following problem:

I computed Area and Perimeter of the polygons and based on these parameters the Compactness Ratio with the following formula:
[Perimeter] / (2 * 3.14159 * Sqr ([Area] / 3.14159))

As I did in other analyses before I wanted to filter the polygons with the lowest compactness (with the highest values of the Compactness Ratio). But due to the fragmented data structure of another layer the polygons are strewn with "holes" (see the attached screenshot, too). This results in rather high values of the Compactness Ratio and makes it really difficult to judge about the fitness of the polygons.

In fact I have to figure out those polygons that allow to build plants in the size of 1500m * 1300m. Does anyone have an idea of how to do that without measuring every single polygon?

Thanks a lot in advance,
Matthias
Tags (3)
1 Solution

Accepted Solutions
by
Esri Regular Contributor
Have a look of my answer to Lisa in this thread:

She wants to fit a square, instead of rectangle. You can use 1/2 of the diagonal length in the process to find candidates.
9 Replies
by
Esri Regular Contributor
Have a look of my answer to Lisa in this thread:

She wants to fit a square, instead of rectangle. You can use 1/2 of the diagonal length in the process to find candidates.
New Contributor III
Dear Dan,

thanks a lot for this hint. The idea is great:

1)   Negative Buffering with the half of the diagonal of my rectangle.
2)   Spatial Join with input polygons and output polygons.

But: It didn't work with the value of half of the diagonal of my rectangle although there obviously would be any more space inside of the input polygons - at least in some of them.

By trying with increasing negative buffering distances I found out that it worked until -370 m (I would need -992 m). I got a message that some of the features have been skipped "because a negative or very small distance resulted in no geometry". That wouldn't be a problem as long as I can push it to the limit. But still there are considerable output polygons in 20 of the input polygons but I can't buffer with bigger distance values.

Why? Is there any reason or idea about how to outsmart ArcMap?

Thanks,
Matthias
by
Esri Regular Contributor
When a polygon has holes, the negative buffer will be generated between the outer ring and borders of the holes. If none of these spaces is wide enough for the buffer, the polygon will be skipped. If you wish the holes to be "ignored", you can use the  Eliminate Polygon Part tool to "fill" some or all of them before running Buffer.
http://resources.arcgis.com/en/help/main/10.1/index.html#//00170000005q000000

Is that what you wish to do?
New Contributor III
Dear Dan,

that is not the point. Some of the polygons in fact have not enough space due to the holes inside. They are skipped and that is OK. But some obviously would have more space (see screenshot). And I wonder why it doesn't work with more than just -370 m buffering distance.

Regards,
Matthias
by
Esri Regular Contributor
Are the red buffers created by -370m? What happens when you use -500m?  Is your data in meters?

Can you share your data?
New Contributor III
Yes, the red buffers are created with -370 m. When using -375 m or more I get the mesage that it doesn't work. The data is in meters. Please find the two shapefiles attached.

Thanks,
Matthias
by
Esri Regular Contributor
Your data is not in meters, but in decimal degrees. Try to project your data and then try Buffer again.

I tried Africa_Equidistant_Conic projection; I can buffer with -700m and get 4 polygons in output.
All polygons are skipped using -900m; not enough room.
New Contributor III
Dear Dan,

In fact, you are right. Stupid mistake of me. Now I projected the data and then it worked.

Thank you anyway!
Matthias
by
Esri Regular Contributor
No problem, Matthias. I am glad it worked out for you.
Have fun geo-processing!