[R-br] Erro função lme

Éder Comunello comunello.eder em gmail.com
Sexta Janeiro 29 10:03:07 BRST 2016


Victor, bom dia

Dei uma olhada rápida no Google e esse problema parece ser bem documentado.
Selecionei dois links para você dar uma olhada, o primeiro me parece ter
uma boa explicação e o segundo uma boa solução (lme4).
"https://stat.ethz.ch/pipermail/r-sig-mixed-models/2009q3/002636.html"
"
http://stackoverflow.com/questions/26449969/backward-selection-in-lme-singularity-in-backsolve-occured
"

Espero que ajude,

​
================================================
Éder Comunello
PhD Student in Agricultural Systems Engineering (USP/Esalq)
Brazilian Agricultural Research Corporation (Embrapa)
Dourados, MS, Brazil [22 16.5'S, 54 49.0'W]




Em 28 de janeiro de 2016 16:52, Victor Fioreze <victorvetzoo em hotmail.com>
escreveu:

> Olá retorno aqui com a minha duvida anterior:
>
> Estou rodando um script com a função lme do pacote nlme e esta dando um
> erro.
> Após eu rodar o modelo usando lme aparece:
>
> Erro em MEEM(object, conLin, control$niterEM) : Singularity in backsolve
> at level 0, block 1
>
> Revisei os dados e o script e esta tudo ok.
> O que pode estar acontecendo?
>
> Att,
>
> pois não sei responder no fórum....
>
> O arquivoa ser carregado esta no anexo
>
> Segue o meu comando para que alguém possa tentar
> > cruza <- read.table("C:/Users/Victor/Dropbox/projeto
> bezerras/Dados/cernelha.txt", header=T,)
> > cruza
>    Bezerro Tratamento   ACd  Data    DN   DJN  Sexo  ACD Idade
> 1      423          1 81.33 41937 41930 41948 FALSE 81.0    67
> 2      423          1 81.59 41944 41930 41948 FALSE 81.0    74
> 3      423          1 82.07 41951 41930 41948 FALSE 81.0    81
> 4      423          1 83.62 41958 41930 41948 FALSE 81.0    88
> 5      423          1 83.72 41965 41930 41948 FALSE 81.0    90
> 6      425          1 81.38 41937 41930 41962 FALSE 81.0    67
> 7      425          1 81.95 41944 41930 41962 FALSE 81.0    74
> 8      425          1 82.40 41951 41930 41962 FALSE 81.0    81
> 9      425          1 83.00 41958 41930 41962 FALSE 81.0    88
> 10     425          1 83.07 41965 41930 41962 FALSE 81.0    90
> 11     426          1 81.09 41937 41930 41968 FALSE 81.0    67
> 12     426          1 81.69 41944 41930 41968 FALSE 81.0    74
> 13     426          1 82.29 41951 41930 41968 FALSE 81.0    81
> 14     426          1 83.20 41958 41930 41968 FALSE 81.0    88
> 15     426          1 83.28 41965 41930 41968 FALSE 81.0    90
> 16     428          1 81.69 41937 41930 41980 FALSE 81.0    67
> 17     428          1 82.32 41944 41930 41980 FALSE 81.0    74
> 18     428          1 83.34 41951 41930 41980 FALSE 81.0    81
> 19     428          1 83.65 41958 41930 41980 FALSE 81.0    88
> 20     428          1 83.74 41965 41930 41980 FALSE 81.0    90
> 21     421          2 82.72 41937 41930 41931 FALSE 79.5    67
> 22     421          2 84.07 41944 41930 41931 FALSE 79.5    74
> 23     421          2 84.98 41951 41930 41931 FALSE 79.5    81
> 24     421          2 85.77 41958 41930 41931 FALSE 79.5    88
> 25     421          2 85.99 41965 41930 41931 FALSE 79.5    90
> 26     429          2 81.29 41937 41930 41989 FALSE 81.0    67
> 27     429          2 83.21 41944 41930 41989 FALSE 81.0    74
> 28     429          2 84.47 41951 41930 41989 FALSE 81.0    81
> 29     429          2 85.75 41958 41930 41989 FALSE 81.0    88
> 30     429          2 85.91 41965 41930 41989 FALSE 81.0    90
> 31     430          2 85.61 41937 41930 42007 FALSE 83.5    67
> 32     430          2 87.20 41944 41930 42007 FALSE 83.5    74
> 33     430          2 87.53 41951 41930 42007 FALSE 83.5    81
> 34     430          2 87.87 41958 41930 42007 FALSE 83.5    88
> 35     430          2 88.02 41965 41930 42007 FALSE 83.5    90
> 36     438          2 90.66 41937 41930 42063 FALSE 89.0    67
> 37     438          2 91.99 41944 41930 42063 FALSE 89.0    74
> 38     438          2 92.80 41951 41930 42063 FALSE 89.0    81
> 39     438          2 94.07 41958 41930 42063 FALSE 89.0    88
> 40     438          2 94.25 41965 41930 42063 FALSE 89.0    90
> 41     422          3 83.43 41937 41930 41933 FALSE 83.0    67
> 42     422          3 84.17 41944 41930 41933 FALSE 83.0    74
> 43     422          3 84.65 41951 41930 41933 FALSE 83.0    81
> 44     422          3 84.91 41958 41930 41933 FALSE 83.0    88
> 45     422          3 84.97 41965 41930 41933 FALSE 83.0    90
> 46     427          3 79.06 41937 41930 41978 FALSE 78.4    67
> 47     427          3 79.86 41944 41930 41978 FALSE 78.4    74
> 48     427          3 80.65 41951 41930 41978 FALSE 78.4    81
> 49     427          3 81.66 41958 41930 41978 FALSE 78.4    88
> 50     427          3 81.78 41965 41930 41978 FALSE 78.4    90
> 51     432          3 82.79 41937 41930 42018 FALSE 80.2    67
> 52     432          3 83.80 41944 41930 42018 FALSE 80.2    74
> 53     432          3 84.64 41951 41930 42018 FALSE 80.2    81
> 54     432          3 85.60 41958 41930 42018 FALSE 80.2    88
> 55     432          3 85.79 41965 41930 42018 FALSE 80.2    90
> 56     435          3 89.95 41937 41930 42025 FALSE 88.5    67
> 57     435          3 90.69 41944 41930 42025 FALSE 88.5    74
> 58     435          3 91.34 41951 41930 42025 FALSE 88.5    81
> 59     435          3 92.30 41958 41930 42025 FALSE 88.5    88
> 60     435          3 92.43 41965 41930 42025 FALSE 88.5    90
> 61     424          4 82.57 41937 41930 41954 FALSE 82.5    67
> 62     424          4 83.02 41944 41930 41954 FALSE 82.5    74
> 63     424          4 83.17 41951 41930 41954 FALSE 82.5    81
> 64     424          4 83.33 41958 41930 41954 FALSE 82.5    88
> 65     424          4 83.35 41965 41930 41954 FALSE 82.5    90
> 66     431          4 82.99 41937 41930 42013 FALSE 81.0    67
> 67     431          4 85.85 41944 41930 42013 FALSE 81.0    74
> 68     431          4 87.54 41951 41930 42013 FALSE 81.0    81
> 69     431          4 89.33 41958 41930 42013 FALSE 81.0    88
> 70     431          4 89.62 41965 41930 42013 FALSE 81.0    90
> 71     433          4 84.94 41937 41930 42020 FALSE 81.4    67
> 72     433          4 86.87 41944 41930 42020 FALSE 81.4    74
> 73     433          4 88.33 41951 41930 42020 FALSE 81.4    81
> 74     433          4 88.99 41958 41930 42020 FALSE 81.4    88
> 75     433          4 89.25 41965 41930 42020 FALSE 81.4    90
> 76     436          4 87.64 41937 41930 42036 FALSE 87.3    67
> 77     436          4 88.81 41944 41930 42036 FALSE 87.3    74
> 78     436          4 89.17 41951 41930 42036 FALSE 87.3    81
> 79     436          4 90.48 41958 41930 42036 FALSE 87.3    88
> 80     436          4 90.58 41965 41930 42036 FALSE 87.3    90
>
> str(cruza)
> 'data.frame':   80 obs. of  9 variables:
>  $ Bezerro   : int  423 423 423 423 423 425 425 425 425 425 ...
>  $ Tratamento: int  1 1 1 1 1 1 1 1 1 1 ...
>  $ ACd       : num  81.3 81.6 82.1 83.6 83.7 ...
>  $ Data      : int  41937 41944 41951 41958 41965 41937 41944 41951 41958
> 41965 ...
>  $ DN        : int  41930 41930 41930 41930 41930 41930 41930 41930 41930
> 41930 ...
>  $ DJN       : int  41948 41948 41948 41948 41948 41962 41962 41962 41962
> 41962 ...
>  $ Sexo      : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
>  $ ACD       : num  81 81 81 81 81 81 81 81 81 81 ...
>  $ Idade     : int  67 74 81 88 90 67 74 81 88 90 ...
>
> library(nlme)
> > attach(cruza)
> > fTrat<-as.factor(Tratamento)
> > class(fTrat)
> [1] "factor"
> > class(ACd)
> [1] "numeric"
> > class(ACD)
> [1] "numeric"
> > class(Data)
> [1] "integer"
> > Data<-as.factor(Data)
> > simcomp.ran <- lme(ACd ~ DJN + ACD + fTrat + Data + fTrat:Data +
> ACD*fTrat + fTrat*DJN,
> + random= ~1 | Bezerro/fTrat)
> Erro em MEEM(object, conLin, control$niterEM) :
>   Singularity in backsolve at level 0, block 1
>
>
> Muito obrigado pela ajuda e desculpe não saber responder no fórum.
>
> Abraço a todos!
>
> Victor Ionatan Fioreze
> Médico Veterinário e Zootecnista
> Mestrando em Nutrição de Ruminantes, PPGZ-UFPel
> Bolsista Capes
> Integrante GERUMEN
> (53) 81032887
> lattes: http://lattes.cnpq.br/5026587958713372
>
>
> _______________________________________________
> R-br mailing list
> R-br em listas.c3sl.ufpr.br
> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça
> código mínimo reproduzível.
>
-------------- Próxima Parte ----------
Um anexo em HTML foi limpo...
URL: <http://listas.inf.ufpr.br/pipermail/r-br/attachments/20160129/afc06f02/attachment.html>


Mais detalhes sobre a lista de discussão R-br