如何向 ggplot 图例添加不同类型的元素？

Question

在底部发布的代码中，我尝试向现有图例添加一个带有标签“置信区间”的灰色条。有人可以帮我将其添加到图例中吗？如果这可以添加到现有的传说中，那就太好了；如果只为这个项目添加一个新的图例更清晰，那也可以。作为旁注，对于渲染的简单绘图，我的 ggplot 代码看起来相当长且混乱，也许这就是 ggplot 代码总是最终的结果。我想知道 ggplot 与基本 R 图相比是否值得。

这张图片显示了我正在尝试做的事情：

代码：

library(dplyr)
library(ggplot2)
library(MASS)
library(survival)

lung1 <- lung %>% 
  mutate(time1 = ifelse(time >= 500, 500, time)) %>% 
  mutate(status1 = ifelse(status == 2 & time >= 500, 1, status))

weibCurve <- function(time, survregCoefs) {exp(-(time/exp(survregCoefs[1]))^exp(-survregCoefs[2]))}

fit1 <- survreg(Surv(time1, status1) ~ 1, data = lung1)

lung1.survfit <- survfit(Surv(time1, status1) ~ 1, data = lung1)
lung1.df <- data.frame(time = lung1.survfit$time, 
                       survival = lung1.survfit$surv, 
                       upper_95 = lung1.survfit$upper, 
                       lower_95 = lung1.survfit$lower)

lung1.df %>%
  ggplot(aes(x = time, y = survival)) +
  geom_ribbon(aes(ymin = lower_95, ymax = upper_95), alpha = 0.2, fill = "grey") +
  geom_line(aes(y = survival, color = "Historical data"), size = 1) +
  scale_x_continuous(limits = c(0, 1500)) +
  scale_y_continuous(limits = c(0, 1), expand = c(0, 0.05)) +
  labs(x = "Time", y = "Fraction surviving", color = NULL) +
  theme_classic() +
  stat_function(fun = weibCurve, args = list(survregCoefs = fit1$icoef), 
                aes(color = "Weibull distribution fit"), size = 1, n = 1000) +
  scale_color_manual(values = c("blue", "red", "grey"), 
                     labels = c("Historical data", "Weibull distribution fit", "Confidence intervals")) +
  labs(color = NULL) +
  theme(legend.position = c(0.95, 0.95), legend.justification = c(1, 1),
        legend.title.align = 0.5, legend.box.spacing = unit(0.3, "lines"), 
        legend.margin = margin(t = 0, r = 0, b = 0, l = 0), legend.title = element_text(size = 12), 
        legend.text = element_text(size = 10))

Answer 1

如果我们将丝带的填充映射到fill美学上，我们可以显示像您的图片一样的图例条目。要控制分配给该填充的颜色，我们可以使用scale_fill_manual，并且可以使用 将填充图例放在颜色图例之后guide_legend(order = ...)。

...
geom_ribbon(aes(ymin = lower_95, ymax = upper_95, fill = "Confidence Interval"), 
            alpha = 0.2) +
scale_fill_manual(values = c("Confidence Interval" = "grey50"), name = NULL) +
guides(color = guide_legend(order = 1), fill  = guide_legend(order = 2)) +
...