使用 ggplot2 绘制拟合的 glm 输出以进行交互

Question

我正在尝试使用 ggplot2 绘制两个不同模型的模型输出。当我想象我的情节时，尽管它们是合适的，但它们看起来完全一样。

两个模型之间的唯一区别是向第二个模型添加了一个变量 "age"（以及包含 "age" 的所有双向交互）。

第一个模型如下：

ab_bi_f= glm(LRS_Bin ~ as.factor(DispersalFate) + MastN + as.factor(DispersalFate)*MastN, data = female, family = binomial(link = logit))

第二种模型如下：

rel_bi_f = glm(LRS_Bin ~ as.factor(DispersalFate) + Age + MastN + as.factor(DispersalFate)*MastN + Age*MastN +as.factor(DispersalFate)*Age, data = female, family = binomial(link = logit))

我用来拟合和可视化第一个模型的代码是：

fitted1<- function (fit) {
    ggplot(data=fit$ab_bi_f, aes(x=MastN, y=LRS_Bin, colour=factor(DispersalFate))) + 
         stat_smooth() + 
         ggtitle("Binary absolute LRS model - female") + 
         labs(x="Total masts in female's lifespan", y="Mean fitted binary absolute LRS", colour="Dispersal type")+
         scale_x_continuous(breaks=c(0,1,2))+
         theme_classic()+annotate("text", x = 0, y = 0.89, label = "a",size=6)+
         coord_cartesian(ylim=c(0.1, 0.9))+
         scale_y_continuous(breaks=c(0.1,0.3,0.5,0.7,0.9))+
         scale_color_manual(values=c("#000000", "#999999"))+
         theme(legend.position=c(0.3,0.85),
         plot.title = element_text(size = 10),
         legend.title=element_text(size=10),
         axis.title=element_text(size=8), 
         legend.key = element_rect(size = 0.1),
         legend.key.size = unit(0.5, "cm"),
         legend.direction="horizontal")
}
plot1<-fitted1(glm(MastN~LRS_Bin))

而且，我用来可视化第二个模型的代码是：

fitted2<- function (fit) {
    ggplot(data=fit$rel_bi_f, aes(x=MastN, y=LRS_Bin, colour=factor(DispersalFate))) + 
         stat_smooth() + 
         ggtitle("Binary absolute LRS model - female") + 
         labs(x="Total masts in female's lifespan", y="Mean fitted binary absolute LRS", colour="Dispersal type")+
         scale_x_continuous(breaks=c(0,1,2))+
         theme_classic()+annotate("text", x = 0, y = 0.89, label = "a",size=6)+
         coord_cartesian(ylim=c(0.1, 0.9))+
         scale_y_continuous(breaks=c(0.1,0.3,0.5,0.7,0.9))+
         scale_color_manual(values=c("#000000", "#999999"))+
         theme(legend.position=c(0.3,0.85),
         plot.title = element_text(size = 10),
         legend.title=element_text(size=10),
         axis.title=element_text(size=8), 
         legend.key = element_rect(size = 0.1),
         legend.key.size = unit(0.5, "cm"),
         legend.direction="horizontal")
}
plot2<-fitted2(glm(MastN~LRS_Bin))

这是结果图：

两个情节看起来一模一样！

我不知道我的代码是否有误，或者这两个模型（尽管第二个模型与第一个模型不同）是否导致相同的输出...

我的数据和代码可以找到here。

Answer 1

您使用的代码存在一些错误。我不确定您是如何获得附加的输出的。但是，如果您只想可视化拟合值。你可以这样做：-

ab_bi_f= glm(LRS_Bin ~ as.factor(DispersalFate) + MastN + as.factor(DispersalFate)*MastN, data = female, family = binomial(link = logit))

rel_bi_f = glm(LRS_Bin ~ as.factor(DispersalFate) + Age + MastN + as.factor(DispersalFate)*MastN + Age*MastN +as.factor(DispersalFate)*Age, data = female, family = binomial(link = logit))

female$fit1 <- predict(ab_bi_f, type = 'response')
female$fit2 <- predict(rel_bi_f, type = 'response')

plot1 <- ggplot(data=female, aes(x=MastN, y=fit1, colour=factor(DispersalFate))) + 
        geom_point() + 
        stat_smooth() + 
        ggtitle("Binary absolute LRS model - female") + 
        labs(x="Total masts in female's lifespan", y="Mean fitted binary absolute LRS", colour="Dispersal type")+
        scale_x_continuous(breaks=c(0,1,2))+
        theme_classic()+annotate("text", x = 0, y = 0.89, label = "a",size=6)+
        coord_cartesian(ylim=c(0.1, 0.9))+
        scale_y_continuous(breaks=c(0.1,0.3,0.5,0.7,0.9))+
        scale_color_manual(values=c("#000000", "#999999"))+
        theme(legend.position=c(0.3,0.85),
        plot.title = element_text(size = 10),
        legend.title=element_text(size=10),
        axis.title=element_text(size=8), 
        legend.key = element_rect(size = 0.1),
        legend.key.size = unit(0.5, "cm"),
        legend.direction="horizontal")

plot2 <- ggplot(data=female, aes(x=MastN, y=fit2, colour=factor(DispersalFate))) + 
        geom_point() + 
        stat_smooth() + 
        ggtitle("Binary absolute LRS model - female") + 
        labs(x="Total masts in female's lifespan", y="Mean fitted binary absolute LRS", colour="Dispersal type")+
        scale_x_continuous(breaks=c(0,1,2))+
        theme_classic()+annotate("text", x = 0, y = 0.89, label = "a",size=6)+
        coord_cartesian(ylim=c(0.1, 0.9))+
        scale_y_continuous(breaks=c(0.1,0.3,0.5,0.7,0.9))+
        scale_color_manual(values=c("#000000", "#999999"))+
        theme(legend.position=c(0.3,0.85),
        plot.title = element_text(size = 10),
        legend.title=element_text(size=10),
        axis.title=element_text(size=8), 
        legend.key = element_rect(size = 0.1),
        legend.key.size = unit(0.5, "cm"),
        legend.direction="horizontal")

grid.arrange(plot1, plot2)

使用 ggplot2 绘制拟合的 glm 输出以进行交互

Plotting fitted glm output using ggplot2 for interactions

plot

interaction

r

ggplot2

data-fitting