Univariate:
    https://www.doc88.com/p-9347835360751.html
    image.png

    Missing Data
    image.png

    主要介绍的是feature plot的用法：
    featurePlot(x, y, plot = if (is.factor(y)) “strip” else “scatter”,
    labels = c(“Feature”, “”), …)

    x为连续变量，或者连续变量的df
    y为factor或者回归中的因变量
    plot：

    • 分类： box, strip, density, pairs or ellipse.
    • 回归：pairs or scatter

    分类：

    1. featurePlot(x = iris[, 1:4], 
    2.             y = iris$Species, 
    3.             plot = "pairs")

    image.png

    1. featurePlot(x = iris[, 1:4], 
    2.             y = iris$Species,
    3.             plot = "density", 
    4.             ## Pass in options to xyplot() to 
    5.             ## make it prettier
    6.             scales = list(x = list(relation="free"), 
    7.                           y = list(relation="free")), 
    8.             adjust = 1.5, 
    9.             pch = "|", 
    10.             layout = c(4, 1), 
    11.             auto.key = list(columns = 3))

    image.png
    回归：

    1. library(mlbench)
    2. data(BostonHousing)
    3. regVar <- c("age", "lstat", "tax")
    4. str(BostonHousing[, regVar])
    6. theme1 <- trellis.par.get()
    7. theme1$plot.symbol$col = rgb(.2, .2, .2, .4)
    8. theme1$plot.symbol$pch = 16
    9. theme1$plot.line$col = rgb(1, 0, 0, .7)
    10. theme1$plot.line$lwd <- 2
    11. trellis.par.set(theme1)
    12. featurePlot(x = BostonHousing[, regVar], 
    13.             y = BostonHousing$medv, 
    14.             plot = "scatter",
    15.             type = c("p", "smooth"),
    16.             span = .5,
    17.             layout = c(3, 1))

    image.png