geom_area在图层之间产生空白区域

Question

我有一个数据框 tbl，其中包含以下数据：

library(tibble)

tbl <- structure(list(
    year = c(2007, 2008, 2008, 2009, 2009, 2010, 2010, 2010, 2011, 2011, 2011, 2011, 2012, 2012, 2012, 2013, 2013, 2013, 2014, 2014, 2014, 2015, 2015, 2015, 2015, 2016, 2016, 2016, 2016, 2017, 2017, 2017, 2018, 2018, 2018, NA), 
    Type = c("RLR", "PLR", "RLR", "PLR", "RLR", "PLR", "RLR", "RR", "PLR", "RLR", "RR", "Supplement", "PLR", "RLR", "RR", "PLR", "RLR", "RR", "PLR", "RLR", "RR", "LR", "PLR", "RLR", "RR", "LR", "Other", "PLR", "RR", "LR", "Other", "RR", "LR", "Other", "RR", "RR"), 
    `n()` = c(4L, 2L, 35L, 14L, 41L, 34L, 37L, 61L, 40L, 21L, 149L, 1L, 38L, 17L, 134L, 41L, 4L, 115L, 23L, 3L, 76L, 1L, 27L, 2L, 78L, 28L, 2L, 4L, 36L, 33L, 5L, 32L, 22L, 3L, 36L, 5L)), 
  row.names = c(NA, -36L), 
  class = c("grouped_df", "tbl_df", "tbl", "data.frame"), 
  groups = structure(list(
    year = c(2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, NA), 
    .rows = list(1L, 2:3, 4:5, 6:8, 9:12, 13:15, 16:18, 19:21, 22:25, 26:29, 30:32, 33:35, 36L)), 
  row.names = c(NA, -13L), 
  class = c("tbl_df", "tbl", "data.frame"), 
  .drop = TRUE))

我想创建一个分层面积图（堆叠图），使用以下代码：

ggplot(tbl,aes(x=year,y=`n()`,fill=Type)) +
geom_area(position="stack") +
theme_light() +
scale_colour_brewer(type="qual",palette = 1) +
ylab("Count")

但是我得到的结果在一些图层之间有奇怪的空白间隙：

我找到了这个问题，但这似乎不是我遇到的问题 - 我的数据中没有负数。

有什么想法可以解决这个问题吗？

这里似乎正在发生：不同时间将从完整列表中添加新的Type并删除旧的Type。当发生这种情况时，这些区域的起点是一个垂直直线，但上方Type的区域会从0斜向上走到这个图层的新起点。这对我来说不像是合理的行为。如何让ggplot将所有颜色都绘制在彼此上方而不是尝试创建带状条纹？我认为这是geom_area的全部意义所在…

- iod

---

如果您的数据框中的空缺以缺失行的形式存在，则需要使用expand.grid和full_join或pivot_wider后跟pivot_longer显式创建缺失行，如此博客文章所述（使用旧名称spread和gather）。 - Paul Rougieux

Answer 1

有一些（年份，类型）的组合缺失了（隐式零）。如果您添加它们（显式零），则没有空白区域。

library(dplyr)     # for %>% operator
library(tidyr)     # for drop_na function

# Your data here....

tbl <- tbl %>%
  # Ungroup or `complete` won't work as expected
  ungroup() %>%
  # There is one NA year
  drop_na() %>%
  # Add all (Type, year) combinations, filling in with 0s where `n()` is not observed
  complete(Type, year, fill = list(`n()` = 0))

ggplot(tbl, aes(x=year,y=`n()`, group=Type, fill=Type)) +
  geom_area(alpha = 0.5) +
  theme_light() +
  scale_colour_brewer(type="qual",palette = 1) +
  ylab("Count")

^{这段内容是由reprex package (v0.2.1)于2019年3月28日创建的。}

- dipetkov