如何在R中将标签作为百分比添加到栅格堆叠的条形图中？

debugcn 发表于 Dev

马里（Iurie Malai）

我已使用包中的likert函数创建了一些堆积的条形图。现在，我想在图表的每个部分中添加以百分比表示的标签，或者更好的是，仅将标签添加到具有足够宽度的那些部分。这该怎么做？我提到我的数据以频率表示，而不是以百分比表示。HHlattice

我的资料：

ssb <- structure(list(`Strongly Disagree` = c(2L, 1L), `Moderate Disagree` = 1:2, 
    `Slightly Disagree` = c(3L, 1L), `Slightly Agree` = c(1L, 
    5L), `Moderate Agree` = 4:5, `Strongly Agree` = c(9L, 6L), 
    Grup = c("Experimental grup", "Control grup")), .Names = c("Strongly Disagree", 
"Moderate Disagree", "Slightly Disagree", "Slightly Agree", "Moderate Agree", 
"Strongly Agree", "Grup"), row.names = c("1", "2"), class = "data.frame")

我的代码：

library(HH)
ppi <- 150
jpeg("ssb_%02d.jpg", width=7*ppi, height=4*ppi, res=ppi)

plot_obj <- likert(Grup ~ . | Grup, data = ssb, as.percent = TRUE, positive.order = TRUE,
    main="", xlab=list(label="Percent", cex=1.1),
    ylab="", ylab.right = list("Subjects per group", cex=1.1),
    scales = list(y = list(relation = "free", labels=""), cex=1.1),
    layout = c(1, 2), auto.key=list(space="bottom", columns=3, title="", cex=1.1))

print(plot_obj)

dev.off()

马里（Iurie Malai）

感谢Deepayan Sarkar，这个问题得以解决：

library(HH)
library(latticeExtra)
ppi <- 150
jpeg("ssb_%02d.jpg", width=7*ppi, height=4*ppi, res=ppi)

plot_obj <- likert(Grup ~ . | Grup, data = ssb, as.percent = TRUE, positive.order = TRUE,
    main="", xlab=list(label="Percent", cex=1.1),
    ylab="", ylab.right = list("Subjects per group", cex=1.1),
    scales = list(y = list(relation = "free", labels=""), cex=1.1),
    layout = c(1, 2), auto.key=list(space="bottom", columns=3, title="", cex=1.1))

plot_obj <- plot_obj +
    layer({
        id = which(x > 0)
        xx = 0.5 * (cumsum(x[id]) + cumsum(c(0, x[id][-length(id)])))
        panel.text(xx, y[id], labels = paste(x[id], "%", sep = ""))
        id = which(x < 0)
        xx = 0.5 * (cumsum(x[id]) + cumsum(c(0, x[id][-length(id)])))
        panel.text(xx, y[id], labels = paste(-x[id], "%", sep = ""))
    })

print(plot_obj)

dev.off()

本文收集自互联网，转载请注明来源。

如有侵权，请联系[email protected] 删除。