<div>Olá Thales!</div>
<div> </div>
<div>Se você adcionar um termo quadrático na expressão de seu modelo Logístico (reg3), seu ajuste vai melhorar em <strong>38%.</strong></div>
<div>E se você adcionar um sexto grau na expressão de seu Modelo Gompertz (mod3), seu ajuste vai melhorar em <strong>12%</strong>.</div>
<div> </div>
<div>Ou seja:</div>
<div> </div>
<div>reg3=gnls(mfs2~((alfa)/(1 + exp(beta - <strong>(gama*daf)^2</strong>))), start=c(alfa=1.5,beta=3,gama=0.015), weights=varExp(form=~daf),correlation=corAR1())<br />mod3=gnls(mfs2~(alfa*<strong>exp(-exp(beta-(gama*daf)))^6</strong>),  start=c(alfa=1.5,beta=3,gama=0.02),weights=varExp(form=~daf),correlation=corAR1())<br /><br /></div>
<div><em>Att.</em></div>
<div><em>André.</em><br /> </div>
<hr style="border-top: 1px solid #ccc;" />
<div>Em 11/09/2012 10:02, <strong>Tales Fernandes < talesest@yahoo.com.br ></strong> escreveu:</div>
<div style="color: #000; background-color: #fff; font-family: times new roman, new york, times, serif; font-size: 12pt;">
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;">Prof. Walmes, muito obrigado.</div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;"> </div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;">Eu não estava compreendendo a diferença entre as situações. </div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;">Lendo a sua descrição acredito que a forma mais correta seria a reg3 (eu já desconfiava disso). Mas o problema é que quando utilizo esta função <span style="font-size: 12pt;">a assíntota superior do modelo fica</span><span style="background-color: transparent;"> sub-estimada. Dessa forma o ajuste fica muito ruim. </span></div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;"><span style="background-color: transparent;"> </span></div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;"><span style="background-color: transparent;">Veja as figuras que são geradas no CRM abaixo.</span></div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;">Saberia me dizer o que está acontecendo ? Ou me indicar a o melhor caminho para me livrar deste problema ?</div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;"> </div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 16px; color: #000000; background-color: transparent; font-style: normal;">Segue o CMR:</div>
<div style="background-color: transparent;">
<div style="background-color: transparent;">rm(list=ls())</div>
<div style="background-color: transparent;">library(car)</div>
<div style="background-color: transparent;">library(nlme)</div>
<div style="background-color: transparent;">library(qpcR) # para calcular o R2</div>
<div style="background-color: transparent;"> </div>
</div>
<div>daf=c(96,111,126,141,158,174,189,204,219,235,250,265,279,293)</div>
<div>mfs2=c(0.014,0.027,0.033,0.189,0.439,0.602,0.653,0.671,0.775,0.814,0.911,1.012,1.059,1.005)</div>
<div> </div>
<div style="color: #000000; font-size: 16px; font-family: 'times new roman', 'new york', times, serif; background-color: transparent; font-style: normal;"># Logístico</div>
<div>reg1=gnls(mfs2~((alfa)/(1 + exp(beta - gama*daf))), start=c(alfa=1.5,beta=3,gama=0.015), weights=varExp(form=~daf),correlation=corAR1(form=~daf))</div>
<div>reg3=gnls(mfs2~((alfa)/(1 + exp(beta - gama*daf))), start=c(alfa=1.5,beta=3,gama=0.015), weights=varExp(form=~daf),correlation=corAR1())</div>
<div> </div>
<div>Rsq(reg1)</div>
<div>Rsq(reg3)</div>
<div> </div>
<div>plot(daf,mfs2, main="Modelo Logístico")</div>
<div>lines(daf,fitted(reg1),col="green")</div>
<div>lines(daf,fitted(reg3),col="red")</div>
<div> </div>
<div># Gompertz</div>
<div>mod1=gnls(mfs2~(alfa*exp(-exp(beta-(gama*daf)))),  start=c(alfa=1.5,beta=3,gama=0.02),weights=varExp(form=~daf),correlation=corAR1(form=~daf))</div>
<div>mod3=gnls(mfs2~(alfa*exp(-exp(beta-(gama*daf)))),  start=c(alfa=1.5,beta=3,gama=0.02),weights=varExp(form=~daf),correlation=corAR1())</div>
<div> </div>
<div>Rsq(mod1)</div>
<div>Rsq(mod3)</div>
<div> </div>
<div>plot(daf,mfs2, main="Modelo Gompertz")</div>
<div>lines(daf,fitted(mod1),col="green")</div>
<div>lines(daf,fitted(mod3),col="red")</div>
<div> </div>
<div style="color: #000000; font-size: 16px; font-family: 'times new roman', 'new york', times, serif; background-color: transparent; font-style: normal;">Atenciosamente.</div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;"> </div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;"> </div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt; text-align: center;"><span style="text-decoration: underline;">Tales Jesus Fernandes</span><br /><span style="text-decoration: underline;">Doutorando em Estatística UFLA</span><br /><span style="text-decoration: underline;">Universidade Federal de Lavras</span></div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;"> </div>
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;">
<div style="font-family: 'times new roman', 'new york', times, serif; font-size: 12pt;">
<div dir="ltr"> </div>
<div id="yiv988273944"> </div>
</div>
</div>
</div>
<div> </div>