R stm 封装图决定文本的一侧

Question

stm教程第19页，图6：主题流行率对比的图形显示

https://cran.r-project.org/web/packages/stm/vignettes/stmVignette.pdf

任何人都知道如何根据点的一侧绘制图形的文本（即“奥巴马”和“莎拉·佩林”位于点的左侧，而“布什总统”位于点的右侧）点）？

Answer 1

estimateEffect

函数的

plot

方法不允许自定义标签定位。但是，您可以将绘图保存到对象并使用元素创建自定义绘图。您还需要添加“omit.plot=TRUE”。

p <- plot(prep, covariate = "rating", topics = c(3, 7, 20),
      model = poliblogPrevFit, method = "difference",
      cov.value1 = "Liberal", cov.value2 = "Conservative",
      xlab = "More Conservative ... More Liberal",
      main = "Effect of Liberal vs. Conservative",
      xlim = c(-.1, .1), labeltype = "custom",
      custom.labels = c('Obama', 'Sarah Palin','Bush Presidency'),
      omit.plot=TRUE)  # Add this

p

是一个列表，其中包含创建相同绘图所需的所有数据。

plot(0, 0, col = "white", 
     ylim=c(0,4), 
     ylab="", yaxt = "n",
     xlim = c(-.1, .1), 
     xlab = "More Conservative ... More Liberal",
     main = "Effect of Liberal vs. Conservative")

it <- length(p$topics)

for (i in 1:length(p$topics)) {
  points(p$means[[i]], it, pch = 16)
  lines(c(p$cis[[i]][1], p$cis[[i]][2]), c(it, it))
  axis(2, at = it, labels = stringr::str_wrap(p$labels[i], width = 25), 
       las = 1, cex = 0.25, tick = FALSE, 
       pos = p$cis[[i]][1])
  it = it - 1
}

这段代码直接来自stm包的

plotDifference

函数。请注意如何使用

axis

函数绘制刻度标签（不带轴）来完成标签。

axis(2, ...)

表示左轴。

我不确定为什么结果与图 6 不同。无论如何，要修改标签的位置，您必须告诉

axis

函数在哪一侧绘制轴（带有相应的刻度标签）。一种方法是先创建值，然后修改轴函数，如下所示：

POS <- c(1, 1, 2)  # The actual position (uses ci - lower [1] or upper [2])
AXIS <- c(2, 2, 4) # which axis (2=left, 4=right)

这将告诉 R 将奥巴马和佩林的标签放在左侧（像以前一样），但将布什的标签放在右侧。现在再次运行上面的代码，除了显示的修改之外：

plot(0, 0, col = "white", 
     ylim=c(0,4), 
     ylab="", yaxt = "n",
     xlim = c(-.1, .1), 
     xlab = "More Conservative ... More Liberal",
     main = "Effect of Liberal vs. Conservative")

lines(c(0, 0), c(0, length(p$topics) + 2), lty = 2)

it <- length(p$topics)

for (i in 1:length(p$topics)) {
  points(p$means[[i]], it, pch = 16)
  lines(c(p$cis[[i]][1], p$cis[[i]][2]), c(it, it))

  # Modify the following line
  axis(AXIS[i], at = it, labels = stringr::str_wrap(p$labels[i], width = 25), 
       las = 1, cex = 0.25, tick = FALSE, 
       pos = p$cis[[i]][POS])
  it = it - 1
}

---