| 1 |
<h2>DESCRIPTION</h2> |
| 2 |
|
| 3 |
<em>v.kernel</em> generates a raster density map from vector points |
| 4 |
data using a moving kernel. Available <a href="https://en.wikipedia.org/wiki/Kernel_(statistics)#Kernel_functions_in_common_use">kernel |
| 5 |
density functions</a> are <em>uniform, triangular, epanechnikov, |
| 6 |
quartic, triweight, gaussian, cosine</em>. The default function is <em>gaussian</em>. |
| 7 |
|
| 8 |
<p>The module can also generate a vector density map on a vector network. |
| 9 |
Conventional kernel functions produce biased estimates by overestimating |
| 10 |
the densities around network nodes, whereas the equal split method of |
| 11 |
Okabe et al. (2009) produces unbiased density estimates. The equal split |
| 12 |
method uses the kernel function selected with the <b>kernel</b> option |
| 13 |
and can be enabled with <b>node=split</b>. |
| 14 |
|
| 15 |
<h2>NOTES</h2> |
| 16 |
|
| 17 |
The <b>multiplier</b> option is needed to overcome the limitation that |
| 18 |
the resulting density in case of a vector map output is stored as category |
| 19 |
(integer). The density result stored as category may be multiplied by this number. |
| 20 |
<p> |
| 21 |
For the <em>gaussian</em> kernel, standard deviation for the |
| 22 |
<a href="https://en.wikipedia.org/wiki/Kernel_(statistics)#Kernel_functions_in_common_use">gaussian function</a> |
| 23 |
is set to 1/4 of the radius. |
| 24 |
<p> |
| 25 |
With the <b>-o</b> flag (experimental) the command tries to calculate an |
| 26 |
optimal radius. The value of <em>radius</em> is taken |
| 27 |
as maximum value. The radius is calculated based on the gaussian function, |
| 28 |
using ALL points, not just those in the current region. |
| 29 |
|
| 30 |
<h2>EXAMPLES</h2> |
| 31 |
|
| 32 |
Compute density of points (using vector map of schools from North Carolina sample dataset): |
| 33 |
<div class="code"><pre> |
| 34 |
g.region region=wake_30m |
| 35 |
v.kernel input=schools_wake output=schools_density radius=5000 multiplier=1000000 |
| 36 |
r.colors map=schools_density color=bcyr |
| 37 |
</pre></div> |
| 38 |
|
| 39 |
<center> |
| 40 |
<img src="v_kernel.png" alt="Density of schools" border="0"><br> |
| 41 |
School density |
| 42 |
</center> |
| 43 |
|
| 44 |
<h2>KNOWN ISSUES</h2> |
| 45 |
|
| 46 |
The module only considers the presence of points, but not |
| 47 |
(yet) any attribute values. |
| 48 |
|
| 49 |
<h2>REFERENCES</h2> |
| 50 |
|
| 51 |
<ul> |
| 52 |
<li>Okabe, A., Satoh, T., Sugihara, K. (2009). <i>A kernel density estimation |
| 53 |
method for networks, its computational method and a GIS-based tool</i>. |
| 54 |
<b>International Journal of Geographical Information Science</b>, Vol 23(1), |
| 55 |
pp. 7-32.<br> |
| 56 |
DOI: <a href="https://doi.org/10.1080/13658810802475491">10.1080/13658810802475491</a></li> |
| 57 |
</ul> |
| 58 |
|
| 59 |
<h2>SEE ALSO</h2> |
| 60 |
|
| 61 |
<em><a href="v.surf.rst.html">v.surf.rst</a></em> |
| 62 |
|
| 63 |
<p> |
| 64 |
Overview: <a href="https://grasswiki.osgeo.org/wiki/Interpolation">Interpolation and Resampling</a> in GRASS |
| 65 |
|
| 66 |
<h2>AUTHORS</h2> |
| 67 |
|
| 68 |
Stefano Menegon, <a href="http://mpa.itc.it/">ITC-irst</a>, Trento, Italy<br> |
| 69 |
Radim Blazek (additional kernel density functions and network part) |