car::scatter3d in R - 更好地标记轴

问题描述

我使用的是 scatter3d，而 3 个轴只有两个端点值.我怎样才能像普通的 plot() 函数一样在整个轴上获取标签?

I'm using scatter3d and the 3 axes just have two endpoint values. how can I get labels throughout the entire axis, just like the normal plot() function does?

推荐答案

哦.我把它当作一个挑战.

Oh well. I took it as a challenge.

显然你需要:

require(rgl)
require(car)
require(mgcv) # for the example

复制 car:::scatter3d.default 代码并将其粘贴回去，将其分配给 scatter3d.default.

Copy the car:::scatter3d.default code and paste it back, assigning it to scatter3d.default.

在 scatter3d.default 的代码中尽早添加这些行:

Add these lines early in the code for scatter3d.default:

 showLabels3d <- car:::showLabels3d
 nice <- car:::nice
  # since you will be losing their connection to the unexposed fns in car

然后在第二个 if(axis.scales){ ...} 后面的代码块中，替换此代码:

Then in the code block following the second if(axis.scales){ ...}, substitute this code:

 if (axis.scales) {
   x.labels <-  seq(lab.min.x, lab.max.x,
                       by=diff(range(lab.min.x, lab.max.x))/4)
   x.at <- seq(min.x, max.x, by=nice(diff(range(min.x, max.x))/4))
      rgl.texts(x.at, -0.05, 0, x.labels, col = axis.col[1])

   z.labels <-  seq(lab.min.z, lab.max.z,
                       by=diff(range(lab.min.z, lab.max.z))/4)
   z.at <- seq(min.z, max.z, by=diff(range(min.z, max.z))/4)
      rgl.texts(0, -0.1, z.at, z.labels, col = axis.col[3])

   y.labels <-  seq(lab.min.y, lab.max.y,
                       by=diff(range(lab.min.y, lab.max.y))/4)
   y.at <- seq(min.y, max.y, by=diff(range(min.y, max.y))/4)
      rgl.texts(-0.05, y.at, -0.05, y.labels, col = axis.col[2])
                }

(您可能需要替换 scatter3d.formula 的代码，这样就不会在 car NAMESPACE 中查找常规调度的 scatter> 方法.我只是将 car::scatter3d.formula 中的 scatter3d 调用替换为scatter3d.default"，以便解释器首先查看新定义的函数.)

(You may need to replace the code for scatter3d.formula so that doesn't look in the car NAMESPACE for the routinely dispatched scatter method. I simply replaced the scatter3d call inside car:::scatter3d.formula with "scatter3d.default" so the interpreter would first look at the newly defined function.)

比使用 scatter3d.formula 搞砸更好的方法是使用以下代码将 car 命名空间/环境分配给新函数:

a better method than mucking with scatter3d.formula is to assign the car namespace/environment to the new function with this code:

environment(scatter3d.default) <- environment(car:::scatter3d.formula)

如果你这样做:

scatter3d(prestige ~ income + education, data=Duncan)

你得到这个(用截图程序拍摄)

You get this (taken with a screenshot program)

这篇关于car::scatter3d in R - 更好地标记轴的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！