<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" /><o:p><FONT size=3 face="Times New Roman"> </FONT></o:p><FONT size=3 face="Times New Roman">Olá Walmes. Muitíssimo obrigado pela dica.</FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><o:p><FONT size=3 face="Times New Roman"> </FONT></o:p></DIV>
<P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><FONT size=3 face="Times New Roman">Aproveitando o exemplo, gostaria de saber se é possível alterar a ordem de comparação das categorias dentro de um fator???, ou seja, como alterar a odds <SPAN style="mso-spacerun: yes"> </SPAN>tipo:</FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><FONT size=3 face="Times New Roman">Factor Método</FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><FONT size=3><FONT face="Times New Roman"><SPAN style="mso-spacerun: yes"> </SPAN>método1 vs método2 </FONT></FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><FONT size=3><FONT face="Times New Roman"><SPAN style="mso-spacerun: yes"> </SPAN>método2 vs método1</FONT></FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt" class=MsoNormal><o:p><FONT size=3 face="Times New Roman"> </FONT></o:p></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><FONT size=3><FONT face="Times New Roman">E como fazer um teste de ajuste do modelo via glm, como o qui-quadrado de<o:p></o:p></FONT></FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><FONT size=3 face="Times New Roman">Pearson e o qui-quadrado da razão de verossimilhanças?????</FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><o:p><FONT size=3 face="Times New Roman"> </FONT></o:p></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><FONT size=3 face="Times New Roman">E se teria uma função que eu possa estudar que crie uma tabela de contingência com as percentagens ou freqüências relativas de cada categoria dentro dos fatores??? Ou um material com este tipo de análise.</FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><o:p><FONT size=3 face="Times New Roman"> </FONT></o:p></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><FONT size=3 face="Times New Roman">Obrigado Walmes.</FONT></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><o:p><FONT size=3 face="Times New Roman"> </FONT></o:p></DIV>
<P style="MARGIN: 0cm 0cm 0pt; mso-layout-grid-align: none" class=MsoNormal><FONT size=3 face="Times New Roman">Marcelo C. Mello</FONT></DIV><BR><BR>--- Em <B>sex, 15/4/11, Walmes Zeviani <I><walmeszeviani@gmail.com></I></B> escreveu:<BR>
<BLOCKQUOTE style="BORDER-LEFT: rgb(16,16,255) 2px solid; PADDING-LEFT: 5px; MARGIN-LEFT: 5px"><BR>De: Walmes Zeviani <walmeszeviani@gmail.com><BR>Assunto: Re: [R-br] Interação significativa em regressão logística<BR>Para: r-br@listas.c3sl.ufpr.br<BR>Data: Sexta-feira, 15 de Abril de 2011, 13:52<BR><BR>
<DIV id=yiv739079050><FONT face="trebuchet ms,sans-serif">Marcelo,<BR><BR>Você não precisa, ao verificar interação, partir para o ajuste do modelo separado. Usar toda a amostra em um mesmo modelo confere mais consistência. Você pode ajustar o modelo de interação com a seguinte fórmula<BR><BR><SPAN style="FONT-FAMILY: courier new, monospace">m0 <- glm(prenhez~época/método, ...)</SPAN><BR><BR>que irá obter o mesmo ajuste, pórem com coeficientes diferentes, pois dessa forma estamos declarando o efeito de época e de método dentro de época. Para facilitar a sua vida, você pode usar a função <SPAN style="FONT-FAMILY: courier new, monospace">contrast::contrast()</SPAN> para fazer a comparação de métodos separados por época<BR><BR><SPAN style="FONT-FAMILY: courier new, monospace">levels(época)</SPAN><BR style="FONT-FAMILY: courier new, monospace"><SPAN style="FONT-FAMILY: courier new, monospace">levels(método)</SPAN><BR
 style="FONT-FAMILY: courier new, monospace"><SPAN style="FONT-FAMILY: courier new, monospace">contrast(m0, list(época="epoca1", método="método1"), </SPAN></FONT><FONT style="FONT-FAMILY: courier new, monospace" face="trebuchet ms,sans-serif">list(época="epoca1", método="método2")</FONT><FONT face="trebuchet ms,sans-serif"><SPAN style="FONT-FAMILY: courier new, monospace">)</SPAN><BR style="FONT-FAMILY: courier new, monospace"><BR>À disposição.<BR>Walmes.<BR><BR>À disposição.<BR>Walmes.<BR><BR clear=all></FONT><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">==========================================================================</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">Walmes Marques Zeviani</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">LEG (Laboratório de Estatística e Geoinformação, 25.450418 S,
 49.231759 W)</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">Departamento de Estatística - Universidade Federal do Paraná</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">fone: (+55) 41 3361 3573</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">VoIP: (3361 3600) 1053 1173</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">e-mail: <A href="http://br.mc1206.mail.yahoo.com/mc/compose?to=walmes@ufpr.br" rel=nofollow target=_blank ymailto="mailto:walmes@ufpr.br">walmes@ufpr.br</A></SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">twitter: @walmeszeviani</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">homepage:
 <A href="http://www.leg.ufpr.br/~walmes" rel=nofollow target=_blank>http://www.leg.ufpr.br/~walmes</A></SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">linux user number: 531218</SPAN><BR style="FONT-FAMILY: trebuchet ms, sans-serif"><SPAN style="FONT-FAMILY: trebuchet ms, sans-serif">==========================================================================</SPAN><BR></DIV><BR>-----Anexo incorporado-----<BR><BR>
<DIV class=plainMail>_______________________________________________<BR>R-br mailing list<BR><A href="http://br.mc1206.mail.yahoo.com/mc/compose?to=R-br@listas.c3sl.ufpr.br" ymailto="mailto:R-br@listas.c3sl.ufpr.br">R-br@listas.c3sl.ufpr.br</A><BR><A href="https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br" target=_blank>https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br</A><BR></DIV></BLOCKQUOTE></td></tr></table>