dear grass users,<br><br>i'm working on landsat imagery and i need to do some automatic classification so i'm doing several tests with i.cluster (first, after that i will also try i.cass or i.gensigset). As I wanted to see the impact of the "sample" parameter, i launch a script with several variations and i noticed this result: when the sample parameter is becoming smaller, i.cluster tend to converge to only 1 class ...<br>
I was thinking that with more samples the result will be more accurate but it seems to be the oposit ... with the default value for sample, the result is always correct ...<br><br>Is there something wrong with the fact to use sample=1,1 (or sample=10,10) instead of the default value sample=56,71 ?<br>
for the moment i'm working with landsat 5 or 7 imagery, so in the case of land use with a very reduce location (ex: "small" rivers), having samples only each 56x15m=840m can occult some land use category ...<br>
<br>I have the same error on gras 6.3 or 6.4, grass did'nt notice any error, and I have enought of ram (only 65% used) ... do you have an idea of what is going wrong and what can i try to solve this problem?<br><br>
<br>I join the results of i.cluster with 20 classes and sample=1,1 parameter set, the others have default values => it goes wrong imediately ... and with default sample (56,71), everything works ...<br><br>Paramètres du cluster<br>
Nombre de classes initiales: 20<br>Taille minimum de classe : 17<br>Séparation minimum entre classes: 0.000000<br>Pourcentage de convergence: 98.000000<br> Nombre maximum d'itérations: 30<br>
<br>Intervalle d'échantillonnage en ligne: 1<br>Intervalle d'échantillonnage en colonne: 1<br><br>Taille de l'échantillon : 40439680 points<br><br>[ ...]<br><br>Distribution des classes<br> 1826562 24892 450783 312030 852164<br>
1646111 2323198 2780468 3019891 3214292<br> 3472239 3559417 3246739 2739526 2249348<br> 2013603 1977504 1754731 1225898 1750284<br><br>######## itération 1 ###########<br>2 classes, 0.01% points stables<br>
Distribution des classes<br> 38277534 0 0 0 0<br> 0 0 0 0 0<br> 0 2162146 0 0 0<br> 0 0 0 0 0<br>
<br>######## itération 2 ###########<br>1 classes, 5.35% points stables<br>Distribution des classes<br> 0 0 0 0 0<br> 0 0 0 0 0<br>
0 40439680 0 0 0<br> 0 0 0 0 0<br><br>######## itération 3 ###########<br>1 classes, 100.00% points stables<br>Distribution des classes<br>
0 0 0 0 0<br> 0 0 0 0 0<br> 0 40439680 0 0 0<br> 0 0 0 0 0<br>
<br><br><br><pre>-- <br>Sylvain Maillard<br>Doctorant en Sciences de l'Environnement<br>
Laboratoire Chimie Provence - UMR 6264 / Université de Provence<br>
la Tour du Valat - Centre de recherche pour la conservation des zones humides méditerranéennes<br>Le Sambuc<br>13200 Arles<br>France<br>mail: <a href="mailto:maillard@tourduvalat.org" target="_blank">maillard@tourduvalat.org</a><br>
tél:04.90.97.29.79<br>fax:04.90.97.20.19<br><a href="http://www.tourduvalat.org/" target="_blank">www.tourduvalat.org</a></pre>
<br>