'[R-sig-Geo] How to work with Poisson distribution (count data)?'

[prev in list] [next in list] [prev in thread] [next in thread] 

List:       r-sig-geo
Subject:    [R-sig-Geo] How to work with Poisson distribution (count data)?
From:       Jimmy Neutron <jimmyjmv () hotmail ! com>
Date:       2015-03-30 17:13:54
Message-ID: BLU182-W5816D072E15B54CCFF5434C4F50 () phx ! gbl
[Download RAW message or body]

[Attachment #2 (multipart/alternative)]

Dear comRades:
I realized that I have to work with Poisson, because I have 'count data' with \
geographic reference. Then, it is 'geodata', such as:  xUTMkm   yUTMkm DyF1385 \
450.1202 1011.425   01386 450.4273 1007.219   01387 450.2584 1011.884   01388 \
450.1696 1010.261   01389 450.1718 1009.887   01390 450.6981 1004.379   0... I read \
the geoRglm structure. What does it mean that I have to make a empiric variogram?. \
What I used when I worked with Binomial model was as \
following:a2008.posCEROS.spmod<-list(cov.pars=c(1,20),beta=1.0,cov.model="matern",nugget=0,kappa=0.35,family="binomial",link="logit")
 Then, is it right to write my Poisson model as \
following?:a2008.posCEROS.spmod<-list(cov.pars=c(1,20),beta=1.0,cov.model="matern",nugget=0,kappa=0.35,family="poisson",link="logit")
 Do I have to calculate nugget and kappa by some method (like likelihood) from my \
'count data' or it is merely a theoretical model?. After that, I will generate MCMC \
simulations, such as: a2008.posCEROS.mcmc<-mcmc.control(S.scale=0.582, thin=10) #mcmc \
marc of change monte carlo, S.scalea2008.posCEROS.tune<-glsm.mcmc(a2008.posCEROSbin, \
model=a2008.posCEROS.spmod, mcmc.input=a2008.posCEROS.mcmc) My goal is to predict how \
many DyF do I have in the survey. Thanks in advance for youR help. 		 	   		  

[Attachment #5 (text/html)]

<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>Dear comRades:<div><br></div><div>I realized \
that I have to work with Poisson, because I have 'count data' with geographic \
reference. Then, it is 'geodata', such as:</div><div><br></div><div><div>&nbsp; \
&nbsp; &nbsp; &nbsp;xUTMkm &nbsp; yUTMkm DyF</div><div>1385 450.1202 1011.425 &nbsp; \
0</div><div>1386 450.4273 1007.219 &nbsp; 0</div><div>1387 450.2584 1011.884 &nbsp; \
0</div><div>1388 450.1696 1010.261 &nbsp; 0</div><div>1389 450.1718 1009.887 &nbsp; \
0</div><div>1390 450.6981 1004.379 &nbsp; \
0</div></div><div>...</div><div><br></div><div><span style="font-size: 12pt;">I read \
the geoRglm structure. What does it mean that I have to make a empiric \
variogram?.</span></div><div><span style="font-size: \
12pt;"><br></span></div><div><span style="font-size: 12pt;">What I used when I worked \
with Binomial model was as following:</span></div><div><span style="font-size: \
12pt;">a2008.posCEROS.spmod&lt;-list(cov.pars=c(1,20),beta=1.0,cov.model="matern",nugget=0,kappa=0.35,family="</span>binomial<span \
style="font-size: 12pt;">",link="logit")</span></div><div><span style="font-size: \
12pt;"><br></span></div><div><span style="font-size: 12pt;">Then, is it right to \
write my Poisson model as following?:</span></div><div><span style="font-size: \
12pt;">a2008.posCEROS.spmod&lt;-list(cov.pars=c(1,20),beta=1.0,cov.model="matern",nugget=0,kappa=0.35,family="</span>poisson<span \
style="font-size: 12pt;">",link="logit")</span></div><div><br></div><div>Do I have to \
calculate&nbsp;<span style="font-size: 12pt;">nugget and kappa by some method (like \
likelihood) from my 'count data' or it is merely a theoretical \
model?.</span></div><div><br></div><div>After that, I will generate MCMC simulations, \
such as:</div><div><br></div><div><div>a2008.posCEROS.mcmc&lt;-mcmc.control(S.scale=0.582, \
thin=10) #mcmc marc of change monte carlo, \
S.scale</div><div>a2008.posCEROS.tune&lt;-glsm.mcmc(a2008.posCEROSbin, \
model=a2008.posCEROS.spmod, \
mcmc.input=a2008.posCEROS.mcmc)</div></div><div><br></div><div>My goal is to predict \
how many DyF do I have in the survey.</div><div><br></div><div>Thanks in advance for \
youR help.</div> 		 	   		  </div></body> </html>

["2008TrialsSuccesses.gif" (image/gif)]
["2008ploteo.gif" (image/gif)]

_______________________________________________
R-sig-Geo mailing list
R-sig-Geo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-geo

[prev in list] [next in list] [prev in thread] [next in thread] 