R_3D图(四)

作者：李誉辉  

四川大学在读研究生 

前言

这篇是plot3D包绘图系列之四，前一篇请戳：R_3D图(三)，后续还有两篇连载，大家一起加油！做教程狠费精力的，别忘了点赞和转发。谢谢。

4 scatter2D()与scatter3D() 及text2D()与text3D()

• point3D()是scatter3D()的特殊形式，参数type = "p"。
  

• lines3D()是scatter3D()的特殊形式，参数type = "l"。
  

• point2D()是scatter2D()的特殊形式，参数type = "p"。
  

• lines2D()是scatter2D()的特殊形式，参数type = "l"。
  

• text2D() 是另一种不可替代的函数
  

• text3D() 是另一种不可替代的函数

语法：

scatter3D (x, y, z, ..., colvar = z, phi = 40, theta = 40,          col = NULL, NAcol = "white", breaks = NULL,          colkey = NULL, panel.first = NULL,          clim = NULL, clab = NULL,          bty = "b", CI = NULL, surf = NULL,          add = FALSE, plot = TRUE)text3D (x, y, z, labels, ..., colvar = NULL, phi = 40, theta = 40,       col = NULL, NAcol = "white",  breaks = NULL,       colkey = NULL, panel.first = NULL,       clim = NULL, clab = NULL,       bty = "b", add = FALSE, plot = TRUE)points3D (x, y, z, ...)lines3D (x, y, z, ...)scatter2D (x, y, ..., colvar = NULL,        col = NULL, NAcol = "white", breaks = NULL,        colkey = NULL, clim = NULL, clab = NULL,        CI = NULL, add = FALSE, plot = TRUE)lines2D(x, y, ...)points2D(x, y, ...)text2D (x, y, labels, ..., colvar = NULL,        col = NULL, NAcol = "white", breaks = NULL, colkey = NULL,        clim = NULL, clab = NULL, add = FALSE, plot = TRUE)

参数解释：

• x, y, z,表示点的坐标，为数字向量，他们应该等长度，length(x) = length(y) = length(z) 。
  

• colvar,表示指定要着色的变量，默认NULL，如果指定，则长度应等于(x, y, z)。
  

• theta, phi, 表示指定观察方向。与persp()中一样。
  

• col, 表示指定色板， xxx.col(), 默认NULL，如colvar指定了，则默认红黄蓝的jet.col()颜色。
  如果col = NULL，且colvar未指定，则col默认为黑色。

• NAcol, 表示指定colvar中NA的颜色。
  

• breaks, 表示指定colvar的断点，为数字向量，长度应该比col参数大1个。
  需要增序排列，默认自动增序排列。
  

• colkey, 为逻辑值或NULL(默认), 也可以用列表传递colkey参数。
  当colkey = NULL时，若col参数是一个向量，才会自动添加图例，col参数是一个字符串则不添加图例。
  设定colkey = list(plot = FALSE)则为图例留下空间，但不显示图例。colkey = FALSE则不绘制图例。
  

• CI，为NULL(默认)或列表(包含参数和置信区间数字向量),
  如果为列表，则至少应包含x, y, z(z仅仅用于scatter3D()中), 这些参数应该是2列的矩阵，表示左/右间隔。 其它参数应该是：alen = 0.01, lty = par(“lty”), lwd = par(“lwd”), col = NULL,
  这几个参数设置箭头的长度，线型，宽度和颜色。 如果col = NULL,则使用colvar指定的颜色。
  

• panel.first, 表示指定一种变换函数，常常用于绘制背景网格和三维散点图的平滑处理。 该函数的其中一个参数应该是pmat矩阵变换。见persp3D()中的例子。

• clab, 表示指定图例标题内容，当colkey = NULL或colkey = FALSE时失效。
  默认位置于主标题同一高度，降低高度，使用向量指定，第一个元素为空字符串。
  

• clim, 表示指定colvar显示范围，如果colvar参数被指定了，则超出clim范围的colvar将以NA显示。
  

• bty, 表示指定box的类型，默认仅仅画背景panels,只有当persp()中的box = TRUE时才有效。
  其它与perspbox()函数中一致，bty = c(“b”, “b2”, “f”, “g”, “bl”, “bl2”, “u”, “n”)其中之一。
  

• labels, 表示指定每个散点的的文本标签内容。其长度应等于(x, y, z)。
  

• surf, NULL(默认)或列表传参, 表示增加散点图拟合的曲面，
  拟合曲面参数包括: 必要参数(x, y, z指定曲面),
  可选参数(colvar, col, NAcol, border, facets, lwd, resfac, clim, ltheta, lphi, shade, lighting, fit)
  参数用法与persp()中一样。 默认参数中未指定colvar, 表示默认colvar = z，z的排列应该与scatter3D()中的z一致。
  

• add, 表示是否将该绘图对象加入到已存在的绘图对象中，TRUE相当于增加图层，默认FALSE则新建。

• plot, 表示是否立即绘图，默认TRUE则立即绘图，FALSE则往下传递绘图参数，直到最后一个图层一起绘制。

• …, 表示其它参数，包括公共参数, persp()中的参数，perspbox()中的一些参数。 persp()中的一些参数：xlim, ylim, zlim, xlab, ylab, zlab, main, sub,
  r, d, scale, expand, box, axes, nticks, ticktype。
  同样xlim,ylim, zlim也只限制坐标轴范围，超出该范围的图形仍然会绘制出来，
  使用plotdev()设定图形范围。 perspbox()中的一些参数： col.axis, col.panel, lwd.panel, col.grid, lwd.grid。

• 公共参数：alpha透明度，从0(全透明)到1(不透明)。 lty线型，lwd线宽，
  

• shade和lighting没有任何作用。
  

• type, 表示指定点或线等绘图几何类型。只有text3D不可取代。
  如type = "p"或"b"，指定type后，pch, cex, bg等参数就可以使用了。

4.1 scatter2D()与scatter3D()

4.1.1 全矩阵数据源

library(plot3D)par(bg = "#b3ff99")# par设定背景颜色M  <- mesh(seq(0, 2*pi, length.out = 100),           seq(0, pi, length.out = 100))u  <- M$x ; v  <- M$yx <- cos(u)*sin(v) # 矩阵Hadamard积，对应元素相乘y <- sin(u)*sin(v)z <- cos(v)#scatter3D(x, y, z, pch = ".", col = "magenta", # pch点型可以为字符         cex = 2, colkey = FALSE, # cex设定点的大小         bty = "u", col.panel = NA) # 手动设定背景, col.panel = NA设定透明背景

4.1.2 type类型

library(plot3D)# 编一个数据z <- seq(0, 10, 0.2)x <- cos(z)y <- sin(z)*z# 三维点图与三维文本scatter3D(x, y, z, phi = 0,   #         col = ramp.col(col = c("cyan", "magenta"), n = length(z)),   # 指定颜色色板         pch = 20, cex = 2, ticktype = "detailed",         bty = "u", col.panel = "#ffffcc", col.grid = "white")## 添加图层, 添加一个点在三维空间scatter3D(x = 0, y = 0, z = 0, add = TRUE,         col = "blue",   # 指定颜色色板         colkey = FALSE,         pch = 10, cex = 3, lwd = 3) # pch点型，cex点尺寸，lwd点中线宽## 添加图层， 添加10个大写字母在三维空间text3D(x = cos(1:10), y = (sin(1:10)*(1:10) - 1),      z = 1:10, colkey = FALSE, add = TRUE,      labels = LETTERS[1:10], col = ramp.col(col = c("blue", "red"), n = 10))

# 三维线图，scatter3D(x, y, z, phi = 0, type = "l",  # type = "l"连线         col = ramp.col(col = c("cyan", "magenta"), n = length(z)),   # 指定颜色色板         ticktype = "detailed", lwd = 4,         bty = "u", col.panel = "#ffffcc", col.grid = "white")

# 三维点图与三维线图scatter3D(x, y, z, phi = 0, type = "b",# type = "b"表示both，即点连线         col = ramp.col(col = c("cyan", "magenta"), n = length(z)),   # 指定颜色色板         bty = "u", col.panel = "#ffffcc", col.grid = "white",           ticktype = "detailed", pch = 20,         cex = c(0.5, 1, 1.5))

# 竖线条scatter3D(x, y, z, phi = 0, type = "h",  # type = "h" 点与竖线条         col = ramp.col(col = c("cyan", "magenta"), n = length(z)),   # 指定颜色色板         bty = "u", col.panel = "#ffffcc", col.grid = "white",         ticktype = "detailed")

4.1.3 CI置信区间

CI参数为列表，内有1到3个元素，每个元素都是矩阵，其行数和列数length(x)。
表示在各个坐标轴方向上的置信区间，

library(plot3D)x <- runif(20)  # 20个随机数y <- runif(20)z <- runif(20)## 设定置信区间参数，1个方向的置信区间CI <- list(z = matrix(nrow = length(x), data = rep(0.05, 2 * length(x))))  # 矩阵行数和列数都是length(x),值均为0.05# 设定box类型，bty='g'灰色背景，白色gridscatter3D(x, y, z, theta = 45, phi = 0, bty = "g", CI = CI, col = ramp.col(col = c("cyan",   "magenta"), n = 100, alpha = 0.8), pch = 19, cex = 2, ticktype = "detailed",   xlim = c(0, 1), ylim = c(0, 1), zlim = c(0, 1))# 增加一些散点 点坐标参数x <- runif(20)y <- runif(20)z <- runif(20)## 设定置信区间参数， 2个方向的置信区间CI2 <- list(x = matrix(nrow = length(x), data = rep(0.05, 2 * length(x))), z = matrix(nrow = length(x),   data = rep(0.05, 2 * length(x))))scatter3D(x, y, z, CI = CI2, add = TRUE, col = "orange", pch = 2)

4.1.4 surf添加曲面

surf参数中，可以使用列表传递曲面参数，与persp()中的参数一样。
当传递坐标参数与scatter3D()中无关则相当于添加曲面。

library(plot3D)M <- mesh(1:nrow(volcano), 1:ncol(volcano))# 绘制100个点，编坐标参数N  <- 100xs <- runif(N) * 87ys <- runif(N) * 61zs <- runif(N)*50 + 154# 绘图,scatter3D(xs, ys, zs, theta = 45, ticktype = "detailed", pch = 16,         bty = "f", xlim = c(1, 87), ylim = c(1,61), zlim = c(94, 215), # bty = "f",所有panels透明         surf = list(x = M$x, y = M$y, z = volcano,           NAcol = "grey", shade = 0.1))

library(plot3D)# 构建曲面网格坐标M <- mesh(seq(0, 2*pi, length = 30), (1:30)/100)z <- with (M, sin(x) + y) ## 散点图坐标N <- 30xs <- runif(N) * 2*piys <- runif(N) * 0.3zs <- sin(xs) + ys + rnorm(N)*0.3CI <- list(z = matrix(nrow = length(xs),                      data = rep(0.3, 2*length(xs))),          lwd = 3) # CI列表传参还能传递其它参数# facets = NA,网格面透明，网格线border为黑色scatter3D(xs, ys, zs, ticktype = "detailed", pch = 16,         col = ramp.col(col = c("cyan", "magenta"), n = 30, alpha = 0.8),         xlim = c(0, 2*pi), ylim = c(0, 0.3), zlim = c(-1.5, 1.5),  # 指定坐标轴显示范围         CI = CI, theta = 20, phi = 30, cex = 2,         surf = list(x = M$x, y = M$y, z = z, border = "orange", facets = NA))

4.1.5 坐标数据来自预测值

在with()内绘图

library(plot3D)par(bg = "#ffffcc")with (mtcars, { # with内处理数据并绘图# 线性回归，多元回归fit <- lm(mpg ~ wt + disp)# 构建预测数据wt.pred <- seq(1.5, 5.5, length.out = 30)disp.pred <- seq(71, 472, length.out = 30)xy <- expand.grid(wt = wt.pred, disp = disp.pred) # expand.grid向量构建数据框## 预测mag值，然后转换成矩阵mpg.pred <- matrix (nrow = 30, ncol = 30,                   data = predict(fit, newdata = data.frame(xy),                                  interval = "prediction")                   )# olddata预测fitpoints <- predict(fit)scatter3D(z = mpg, x = wt, y = disp, pch = 18, cex = 2,         theta = 20, phi = 20, ticktype = "detailed",         col = ramp.col(col = c("cyan", "magenta"), n = 30, alpha = 0.8),         xlab = "wt", ylab = "disp", zlab = "mpg",  # 修改坐标轴标题         surf = list(x = wt.pred, y = disp.pred, z = mpg.pred,                       facets = NA, fit = fitpoints), # fit参数增加预测值与真实值之间的连线         main = "mtcars", bty = "u", col.panel = NA) # col.panel = NA则panel透明})

4.1.6 panel.first参数在panel中增加图形

panel.first可以指定绘图对象，也就可以在panels中增加图形。

library(plot3D)par(bg = "#ffffcc")# 第一种方法，type = "h"增加垂线条with(quakes, scatter3D(x = long, y = lat, z = -depth, colvar = mag,    pch = 16, cex = 1.5, xlab = "longitude", ylab = "latitude",    col = ramp.col(col = c("cyan", "magenta"), n = length(mag), alpha = 0.5),    zlab = "depth, km", clab = c("Richter","Magnitude"),    main = "Earthquakes off Fiji", ticktype = "detailed", bty = "u", col.panel = NA, # 透明面板    type = "h", theta = 10, d = 2,    colkey = list(length = 0.5, width = 0.5, cex.clab = 0.75)) # 图例列表传参)# 第2中方法，使用trans3D构建转换空间# 然后使用转换空间绘制二维散点图，再使用panel.first参数将该图add到scatter3D基图中# 绘制panel.first图lens <- length(quakes$mag)panelfirst <- function(pmat) { ## 绘制x-y平面的panel zmin <- min(-quakes$depth) ### trans3D构建转换空间，x,y坐标与第1种方法中一样，z轴仅仅计数，以免重复 XY <- trans3D(quakes$long, quakes$lat,               z = rep(zmin, nrow(quakes)), pmat = pmat) ### 使用转换空间绘制散点图，必须使用add = TRUE才能将图形以panel.first参数传入基图中 scatter2D(XY$x, XY$y, colvar = quakes$mag, pch = ".", # colvar着色变量与基图相同           col = ramp.col(col = c("cyan", "magenta"), n = lens, alpha = 0.5),           cex = 2, add = TRUE, colkey = FALSE) # ## 绘制y-z平面的panel xmin <- min(quakes$long) # x轴仅仅计数 ### trans3D构建转换空间，y,z坐标与第1种方法相同，x轴仅仅计数避免重复 XY <- trans3D(x = rep(xmin, nrow(quakes)), y = quakes$lat,               z = -quakes$depth, pmat = pmat) ### 使用转换空间绘制散点图，必须使用add = TRUE才能将图形以panel.first参数传入基图中 scatter2D(XY$x, XY$y, colvar = quakes$mag, pch = ".", # colvar着色变量与基图相同           col = ramp.col(col = c("cyan", "magenta"), n = lens, alpha = 0.5),           cex = 2, add = TRUE, colkey = FALSE)}# 将panel.first图添加到scatter3D基图中with(quakes, scatter3D(x = long, y = lat, z = -depth, colvar = mag,    pch = 16, cex = 1.5, xlab = "longitude", ylab = "latitude",    col = ramp.col(col = c("cyan", "magenta"), n = length(mag), alpha = 0.5),    zlab = "depth, km", clab = c("Richter","Magnitude"),    main = "Earthquakes off Fiji", ticktype = "detailed", bty = "u", col.panel = NA, # 透明面板    panel.first = panelfirst, theta = 10, d = 2,  # panel.first传递panels绘图对象    colkey = list(length = 0.5, width = 0.5, cex.clab = 0.75)))

4.2 text2D()与text3D()

4.2.1 text3D()添加文字

library(plot3D)par(bg = "#ffffcc")# par设定背景颜色data("USArrests")# 绘制散点图与垂线条scatter3D(USArrests$Murder, USArrests$Assault, USArrests$Rape - 1,         colvar = USArrests$UrbanPop,         col = ramp.col(col = c("cyan", "magenta"), n = length(USArrests$UrbanPop), alpha = 0.5),         colkey = list( length = 0.5, col.clab = "blue", dist = -0.1), # 列表传参,设定图例         type = "h", pch = ".",bty = "u", col.panel = NA)# 添加文字text3D(USArrests$Murder, USArrests$Assault, USArrests$Rape, add = TRUE,      colvar = USArrests$UrbanPop, theta = 60, phi = 20,      col = ramp.col(col = c("cyan", "magenta"), n = length(USArrests$UrbanPop), alpha = 0.5),      xlab = "Murder", ylab = "Assault", zlab = "Rape",      main = "USA arrests", colkey = FALSE,      labels = rownames(USArrests), cex = 0.6,      bty = "u", ticktype = "detailed", d = 2,      clab = c("Urban","Pop"), adj = 0.5, font = 2, col.panel = NA)

4.2.2 坐标轴刻度文本text

library(plot3D)data(VADeaths)hist3D(x = 1:5, y = 1:4, z = VADeaths, expand = 0.5, phi = 20, theta = -72,   shade = 0.2, itheta = 72, iphi = 60, d = 2, xlab = "", ylab = "", zlab = "",   main = "VADeaths", col = ramp.col(col = c("cyan", "green"), n = 3, alpha = 0.5)[2],   border = "magenta", alpha = 0.15, opaque.top = TRUE, bty = "g", space = 0.3,   ticktype = "detailed", cex.axis = 1e-09)text3D(x = 1:5, y = rep(0.5, 5), z = rep(3, 5), labels = rownames(VADeaths),   add = TRUE, adj = 0, col = "magenta", bty = "g")text3D(x = rep(1, 4), y = 1:4, z = rep(0, 4), labels = colnames(VADeaths), add = TRUE,   adj = 1, col = "magenta", bty = "g")

4.3 scatter2D()

4.3.1 type参数

library(plot3D)par(mfrow = c(2, 2), bg = "#ffffcc") # 多图排版，2*2矩阵排列x <- seq(0, 2*pi, length.out = 30)# 默认type为点scatter2D(x, sin(x), colvar = cos(x), pch = 16,         col = ramp.col(col = c("cyan", "magenta"), n = 30, alpha = 0.5),         colkey = list(col.axis = "blue", col.ticks = "green", col.clab = "purple"), # 设定图例参数         ylab = "sin", clab = "cos", cex = 1.5)# type = "l"连线scatter2D(x, sin(x), colvar = cos(x),         col = ramp.col(col = c("cyan", "magenta"), n = 30, alpha = 0.5),         colkey = list(col.axis = "blue", col.ticks = "green", col.clab = "purple"), # 设定图例参数        type = "l", lwd = 4, bty = "g")# type = "b"点和联系scatter2D(x, sin(x), colvar = cos(x),         col = ramp.col(col = c("cyan", "magenta"), n = 30, alpha = 0.5),         colkey = list(col.axis = "blue", col.ticks = "green", col.clab = "purple"), # 设定图例参数         type = "b", lwd = 2, bty = "g")# type = "h"竖线scatter2D(x, sin(x), colvar = cos(x),         col = ramp.col(col = c("cyan", "magenta"), n = 30, alpha = 0.5),         colkey = list(col.axis = "blue", col.ticks = "green", col.clab = "purple"), # 设定图例参数         type = "h", lwd = 4, alpha = 0.5) # 指定透明的和线宽

4.3.2 text2D()添加文本

library(plot3D)data("USArrests")par(bg = "#ffffcc")text2D(x = USArrests$Murder, y = USArrests$Assault + 5, colvar = USArrests$Rape,   col = ramp.col(col = c("cyan", "magenta"), n = length(USArrests$Rape), alpha = 0.5),   xlab = "Murder", ylab = "Assault", clab = "Rape", main = "USA arrests",   labels = rownames(USArrests), cex = 0.6, adj = 0.5, font = 2, colkey = list(plot = FALSE))scatter2D(x = USArrests$Murder, y = USArrests$Assault, colvar = USArrests$Rape,   col = ramp.col(col = c("cyan", "magenta"), n = length(USArrests$Rape), alpha = 0.8),   pch = 16, add = TRUE)

4.3.3 mesh数据源

library(plot3D)par(mfrow = c(1, 2), bg = "#ffffcc")x <- seq(-1, 1, by = 0.1)y <- seq(-2, 2, by = 0.2)grid <- mesh(x, y)z <- with(grid, cos(x) * sin(y))# 第一张图image2D(x = x, y = y, z = z, col = ramp.col(col = c("cyan", "magenta"), n = length(x),   alpha = 0.5))points(grid, col = "blue")  # 使用R自带函数添加图层,设定变量颜色为蓝色# 第2张图scatter2D(grid$x, grid$y, colvar = z, pch = 20, cex = 2, col = ramp.col(col = c("cyan",   "magenta"), n = length(x), alpha = 0.5))

4.3.4 CI置信区间

library(plot3D)par(bg = "#ffffcc")x <- sort(rnorm(10))y <- runif(10)cv <- sqrt(x^2 + y^2)# 创建置信区间列表CI <- list(lwd = 2)  # 置信区间线宽CI$x <- matrix(nrow = length(x), data = c(rep(0.25, 2 * length(x))))  # 给列表增加元素，x方向的置信区间CI2 <- CICI2$y <- matrix(nrow = length(y), data = c(rep(0.25, 2 * length(y))))  # 给列表增加元素，y方向的置信区间scatter2D(x, y, colvar = cv, col = ramp.col(col = c("cyan", "magenta"), n = length(x),   alpha = 0.5), pch = 16, cex = 2, CI = CI)# 更改type = 'b'点连线scatter2D(x, y, colvar = cv, col = ramp.col(col = c("cyan", "magenta"), n = length(x),   alpha = 0.5), pch = 16, cex = 2, CI = CI, type = "b")# 增加y方向的置信区间scatter2D(x, y, colvar = cv, col = ramp.col(col = c("cyan", "magenta"), n = length(x),   alpha = 0.5), pch = 16, cex = 2, CI = CI2, type = "b")

4.3.5 添加绘图对象到地图中

library(plot3D)data(Oxsat)# image2D绘制地图oxlim <- range(Oxsat$val[, , 1], na.rm = TRUE)image2D(z = Oxsat$val[, , 1], x = Oxsat$lon, y = Oxsat$lat, contour = TRUE,   xlab = "longitude", ylab = "latitude", main = "Oxygen saturation", clim = oxlim,   clab = "%")# 数据点坐标创建lon <- c(11.2, 6, 0.9, -4, -8.8)lat <- c(-19.7, -14.45, -9.1, -3.8, -1.5)O2sat <- c(90, 95, 92, 85, 100)  # 数据点着色变量# add to image; use same zrange; avoid adding a color keyscatter2D(x = lon, y = lat, colvar = O2sat, clim = oxlim, pch = 16, add = TRUE,   cex = 2, colkey = FALSE)

4.4 line2D()

与scatter2D()中type = "l"结果一样

library(plot3D)data("sunspot.month")par(bg = "#ffffcc")# 筛选数据sunspot <- data.frame(year = time(sunspot.month),                     anom = sunspot.month - mean(sunspot.month))ff <- 100sunspot$ma <- filter(sunspot$anom, rep(1/ff, ff), sides = 2)# type = "h"绘制垂线条lines2D(sunspot$year, sunspot$anom, colvar = sunspot$anom > 0,       col = ramp.col(col = c("pink", "lightblue"), n = 2, alpha = 0.1), # 正数一种颜色，负数一种颜色       main = "sunspot anomaly", type = "h",       colkey = FALSE, las = 1, xlab = "year", ylab = "")# tyep = "l"绘制连线,注意着色变量不一样lines2D(sunspot$year, sunspot$ma, add = TRUE,       col = ramp.col(col = c("green", "magenta"), n = length(sunspot$ma), alpha = 0.9))  

····

往期精彩：

• R_插值_拟合_回归_样条
  

• R_circlize包_和弦图（一）
  

• R_circlize包_和弦图（二）

• R_ggplot2基础（一）

• R_ggplot2基础（二）

• R_ggplot2基础（三）

• R_ggplot2基础（四）
  

• R_3D图(一)

• R_3D图(二)

····

公众号后台回复关键字即可学习

回复 爬虫            爬虫三大案例实战  
回复 Python       1小时破冰入门

回复 数据挖掘     R语言入门及数据挖掘
回复 人工智能     三个月入门人工智能
回复 数据分析师  数据分析师成长之路 
回复 机器学习      机器学习的商业应用
回复 数据科学      数据科学实战
回复 常用算法      常用数据挖掘算法