格子图：添加每个 x 值具有多个 y 值的多项式回归线

Question

我正在尝试将具有多项式回归模型的回归线添加到我的 xyplot 中，该 xyplot 具有每个 x 值的多个 y 值，这会创建多个 ablines（我认为）。

amp.plot.1 <- structure(list(TP = c(-2, -1, 0, 1, 2, 4, 6, 8, 9, -2, -1, 0, 
1, 2, 5, 7, 8, 9, 4, 6), SID_2 = c("UAB003_1", "UAB003_1", "UAB003_1", 
"UAB003_1", "UAB003_1", "UAB003_1", "UAB003_1", "UAB003_1", "UAB003_1", 
"UAB003_2", "UAB003_2", "UAB003_2", "UAB003_2", "UAB003_2", "UAB003_2", 
"UAB003_2", "UAB003_2", "UAB003_2", "UAB005", "UAB005"), UID = c("UAB003_W1D3", 
"UAB003_W1D6", "UAB003_W2D1", "UAB003_W2D2", "UAB003_W2D3", "UAB003_W2D5", 
"UAB003_W2D7", "UAB003_W3D2", "UAB003_W3D3", "UAB003_W8D3", "UAB003_W8D5", 
"UAB003_W9D3", "UAB003_W9D4", "UAB003_W9D5", "UAB003_W10D1", 
"UAB003_W10D3", "UAB003_W10D6", "UAB003_W10D7", "UAB005_W2D4", 
"UAB005_W2D6"), Species = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), levels = c("Lactobacillus_iners", 
"Lactobacillus_crispatus", "Lactobacillus_jensenii", "Lactobacillus_gasseri", 
"Gardnerella_vaginalis", "Sneathia_sanguinegens", "Atopobium_vaginae", 
"BVAB1", "Prevotella_timonensis", "Prevotella_amnii", "Prevotella_bivia", 
"g_Megasphaera"), class = "factor"), Count = c(5.54644502967252, 
4.88834003849887, 4.31944507354567, 5.47828041152078, 5.59699483640575, 
5.66255777037516, 5.80752511653788, 5.67881819020191, 5.59099252829171, 
3.98942397054398, 4.07929361555486, 4.23489117686723, 3.50636505344502, 
4.47583162261933, 5.23018702562597, 5.14641307286776, 5.18195301770781, 
3.72092821777661, 5.76233696844788, 5.90702924988876)), row.names = c("1", 
"2", "3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13", 
"14", "15", "16", "17", "18", "19", "20"), class = "data.frame")

xyplot(amp.plot.1$Count ~ amp.plot.1$TP | amp.plot.1$Species, 
            aspect = 1:1,
            groups = amp.plot.1$SID_2,
            panel = function(x, y, ...) {
              panel.xyplot(x, y, ...)
              fm = lm(y ~ poly(x, 3))
              panel.lines(x, fitted(fm), col.line = "red")
              panel.text(3, 1.5, labels = sprintf("y = %.3f x³ + %.3f x² + %.3f x + %.3f\nR² = %.2f", 
                                                  coef(fm)[4], coef(fm)[3], coef(fm)[2], coef(fm)[1], summary(fm)$r.squared), 
                                                  cex = 0.7)},
            scales = "free",
            pch = 1, 
            cex = .5,
            xlim = c(-3, 9),
            ylim = c(0.5, NA),
            ylab = list(label = "log10[estimated absolute abundance + 1]", fontsize = 15),
            xlab = list(label = "Day of MET Treatment", fontsize = 15),
            par.strip.text = list(cex = 0.8),
            layout = c(4, 3)
) 

当我用线性回归绘制相同的图时，它工作得很好。当我用 ggplot 做同样的事情时，回归线是正确的。 有人可以帮我找出修复 xyplot 中回归线的正确方法，这样每个方面只有一个吗？ 谢谢！

Answer 1

根据 Deepayan 的说法，“您的方法的问题是您定义的面板函数不处理组。一个简单的解决方法是使用

panel.superpose

”

我认为您还必须在

xyplot

块之外创建模型。

fm = lm(Count ~ poly(TP, 3), data=amp.plot.1)

然后将其添加为

fit

参数以获得所需的曲线。

xyplot(Count ~ TP | Species, data=amp.plot.1, aspect = 1:1, fit=fm,
       groups = SID_2,
       panel = panel.superpose,
       panel.groups = function(x, y, ..., fit, subscripts) {
         panel.xyplot(x, y, ...)
         ypred <- fitted(fit)[subscripts]
         panel.lines(x, ypred, col = "black")
       },
... # the remaining as you provided.