ggproto:如何访问compute_layer()中的非美学参数值？

Question

我在试着做新的衣服和数据。我尝试了来自这个StatChull的格列奈特代码。我的目标是操纵一个外部参数，这不是一个审美价值。是这样的：

stat_custom(data = df, mapping = aes(x = xval, y = val), myparam = myval, geom = "custom")

问题是，当我使用compute_group()创建自定义stat时，我可以获得自定义参数。一旦我将compute_group()更改为compute_layer()，程序就会停止工作。

下面是stat_chull()的工作程序：

StatChull <- ggproto("StatChull", Stat,
                     compute_group = function(self, data, scales, params, na.rm, myparam) {
                       message("My param has value ", myparam)
                       # browser()
                       data[chull(data$x, data$y), , drop = FALSE]
                     },

                     required_aes = c("x", "y")
)

stat_chull <- function(mapping = NULL, data = NULL, geom = "polygon",
                       position = "identity", na.rm = FALSE, myparam = "", show.legend = NA, 
                       inherit.aes = TRUE, ...) {
  layer(
    stat = StatChull, data = data, mapping = mapping, geom = geom, 
    position = position, show.legend = show.legend, inherit.aes = inherit.aes,
    params = list(na.rm = na.rm, myparam = myparam, ...)
  )
}

ggplot(mpg, aes(displ, hwy)) + 
  geom_point() + 
  stat_chull(fill = NA, colour = "black", myparam = "myval")

这将打印在控制台上：

My param has value myval

当我将compute_group()更改为compute_layer()时，此程序会出错

StatChull <- ggproto("StatChull", Stat,
                     compute_layer = function(self, data, scales, params, na.rm, myparam) {
                       message("My param has value ", myparam)
                       # browser()
                       data[chull(data$x, data$y), , drop = FALSE]
                     },

                     required_aes = c("x", "y")
)

stat_chull <- function(mapping = NULL, data = NULL, geom = "polygon",
                       position = "identity", na.rm = FALSE, myparam = "", show.legend = NA, 
                       inherit.aes = TRUE, ...) {
  layer(
    stat = StatChull, data = data, mapping = mapping, geom = geom, 
    position = position, show.legend = show.legend, inherit.aes = inherit.aes,
    params = list(na.rm = na.rm, myparam = myparam, ...)
  )
}

ggplot(mpg, aes(displ, hwy)) + 
  geom_point() + 
  stat_chull(fill = NA, colour = "black", myparam = "myval")

这将打印在控制台上：

Warning: Ignoring unknown parameters: myparam

Error in message("My param has value ", myparam): argument "myparam" is missing, with no default

有人能告诉我如何在compute_layer()中访问参数值吗？

Answer 1

解释

所有geom_*()和stat_*()函数都是围绕layer()的包装器。如果我们检查其中的代码，我们可以看到与Stat关联的参数值是在stat_params中捕获的，这是通过stat$parameters(TRUE)获得的，其中stat引用了特定的Stat* ggproto对象：

> layer
function (geom = NULL, stat = NULL, data = NULL, mapping = NULL, 
    position = NULL, params = list(), inherit.aes = TRUE, check.aes = TRUE, 
    check.param = TRUE, show.legend = NA) 
{
    ... # omitted
    params <- rename_aes(params)
    aes_params <- params[intersect(names(params), geom$aesthetics())]
    geom_params <- params[intersect(names(params), geom$parameters(TRUE))]
    stat_params <- params[intersect(names(params), stat$parameters(TRUE))]
    all <- c(geom$parameters(TRUE), stat$parameters(TRUE), geom$aesthetics())
    extra_param <- setdiff(names(params), all)
    if (check.param && length(extra_param) > 0) {
        warning("Ignoring unknown parameters: ", paste(extra_param, 
            collapse = ", "), call. = FALSE, immediate. = TRUE)
    }
    extra_aes <- setdiff(mapped_aesthetics(mapping), c(geom$aesthetics(), 
        stat$aesthetics()))
    if (check.aes && length(extra_aes) > 0) {
        warning("Ignoring unknown aesthetics: ", paste(extra_aes, 
            collapse = ", "), call. = FALSE, immediate. = TRUE)
    }
    ggproto("LayerInstance", Layer, geom = geom, geom_params = geom_params, 
        stat = stat, stat_params = stat_params, data = data, 
        mapping = mapping, aes_params = aes_params, position = position, 
        inherit.aes = inherit.aes, show.legend = show.legend)
}

StatChull从Stat继承而不更改其参数函数，这将是：

> Stat$parameters
<ggproto method>
  <Wrapper function>
    function (...) 
f(..., self = self)

  <Inner function (f)>
    function (self, extra = FALSE) 
{
    panel_args <- names(ggproto_formals(self$compute_panel))
    group_args <- names(ggproto_formals(self$compute_group))
    args <- if ("..." %in% panel_args) 
        group_args
    else panel_args
    args <- setdiff(args, names(ggproto_formals(Stat$compute_group)))
    if (extra) {
        args <- union(args, self$extra_params)
    }
    args
}

解决方案

从上面可以看出，stat$parameters依赖于相关的Stat*的compute_panel / compute_group函数中列出的函数参数。

StatChull <- ggproto("StatChull", 
                     Stat,
                     compute_layer = function (self, data, params, layout) {
                       message("My param has value ", params$myparam)
                       data[chull(data$x, data$y), , drop = FALSE]
                     },
                     compute_group = function(self, data, scales, na.rm, myparam) {
                       # this function is never triggered, but defined here
                       # in order for myparam to be included in stat_params
                     },
                     required_aes = c("x", "y")
)

# no change to stat_chull

请注意，我们已经为compute_group定义了一个平凡的StatChull函数。此函数从未触发，因为compute_layer直接返回数据集(在Stat的正常情况下，compute_group由compute_panel触发，而后者又由compute_layer触发)，但由于它确实将myparam作为其函数参数之一，所以myparam现在被认为是params中的参数值之一。

(您也可以通过定义一个简单的compute_panel函数来实现同样的结果。)

示范：

ggplot(mpg, aes(displ, hwy)) + 
  geom_point() + 
  stat_chull(fill = NA, colour = "black", myparam = "myval")

​

​

My param has value myval