[AniMov] Understanding the output of Kernel Density Estimate from AniMove

JOHN CHAN ust.johnchan at gmail.com
Tue Feb 26 04:26:55 CET 2019


*This is my first time using this mailing list. Thank you all in advance.*

*I have asked the question in GIS stackExchange:*

I found the AniMove plugin for QGIS to do the Home Range Analysis for
Dolphins.

I load all the points of the sighting, and perform the task using the
plugin. I got the contour as polyline, and heatmap as raster. But I have
trouble understanding the meaning.

No matter what Utilization Distribution value I input in the parameters
box, I get the same contour lines.

And I don't know what is the value in raster means. I tried to check the
source code. It seems that the value is estimated by kernel-density
estimate using Gaussian kernels. And the z value is further modified by Z =
Z.clip(0) * 100.0 / Z.max().

*I have figured out the following:*

I have figured out some of the answers myself. First, the contour is in 10
incremental level. The inner contour is 10% homerange. The outer contour is
90% homerange. It seems that the plugin indeed calculate everything no
matter what utilisation distribution value you plug in. The value in the
raster is a bit tricky. First, it estimates the kernel density, called
value Z. The normalised them by the max value. So, the 95% home range is
actually represents by all raster value >=5.

*Please let me know if I am correct or not.*

Best Regards,
John


More information about the AniMov mailing list