color=date在ggplot2中的控制

Question

在使用以下代码之前/之后，我正在做一个足够简单的行图：

ggplot() +
  geom_line(data = preMax, aes(x=buffer, y=max, color=date, linetype= "dashed")) +
  geom_line(data = postMax, aes(x=buffer, y=max, color=date, linetype= "solid"))

这让我得到了下面的情节：

​

​

有些问题我似乎解决不了。

priority)

• Labelling

• 线型图例是倒的

• 月份的颜色调色板是颠倒的(这是在图的右边的每一行(我知道这会变得很混乱--这些只是诊断图)

虽然我已经为3)找到了很多解决方案，但我无法为我的实现找到语法。

我不知道该为1和2做些什么。

前后数据格式的摘录(请原谅从CSV导入中读取的杂乱无章的标题--“掩码”列在这里是多余的)：

> dput(preMax[sample(1:nrow(preMax), 20),])
structure(list(type = c("Masked", "Masked", "Masked", "Masked", 
"Masked", "Masked", "Masked", "Masked", "Masked", "Masked", "Masked", 
"Masked", "Masked", "Masked", "Masked", "Masked", "Masked", "Masked", 
"Masked", "Masked"), buffer = c(200, 1700, 600, 400, 900, 300, 
1300, 600, 1500, 100, 400, 1700, 2000, 1800, 300, 1000, 100, 
1000, 1900, 1800), max = c(44.864783808451, 45.9712411935412, 
17.5829020162918, 12.4868639407346, 49.6003614709968, 45.2344134295274, 
21.1563882993978, 10.5763618088791, 37.6644106683852, 11.057020709424, 
33.3232497692325, 20.7153387275033, 37.6645113431222, 49.9596451668152, 
11.0570186079038, 45.6029948612003, 21.156362982432, 37.6638859359373, 
20.2727947583503, 21.5957031868673), date = c("2011-07-13", "2011-07-13", 
"2010-02-16", "2010-12-01", "2010-08-11", "2011-07-13", "2010-11-15", 
"2010-01-15", "2011-04-24", "2010-01-15", "2011-10-17", "2010-03-20", 
"2011-04-24", "2010-08-11", "2010-01-15", "2011-07-13", "2010-03-20", 
"2011-04-24", "2010-03-20", "2010-11-15")), row.names = c(NA, 
-20L), class = c("tbl_df", "tbl", "data.frame"))

> dput(postMax[sample(1:nrow(postMax), 20),])
structure(list(type = c("Masked", "Masked", "Masked", "Masked", 
"Masked", "Masked", "Masked", "Masked", "Masked", "Masked", "Masked", 
"Masked", "Masked", "Masked", "Masked", "Masked", "Masked", "Masked", 
"Masked", "Masked"), buffer = c(1000, 500, 100, 200, 600, 1300, 
300, 1100, 400, 1400, 100, 1400, 1900, 1800, 900, 1800, 400, 
1700, 800, 1300), max = c(14.7852734070896, 14.9722784687354, 
15.2029852639159, 43.729031912867, 15.1848393092082, 51.7461939052593, 
24.5462660651035, 14.7020628187789, 51.660784286412, 51.6686452841226, 
37.4059816958464, 31.1296293850412, 16.3389454899921, 14.9904419060676, 
9.31826862882747, 37.5157587930802, 42.542644563124, 37.6102405759581, 
24.2312294619778, 42.7078563816514), date = c("2017-12-04", "2017-12-04", 
"2018-02-22", "2017-08-14", "2018-02-22", "2017-07-13", "2017-03-07", 
"2017-12-04", "2017-07-13", "2017-07-13", "2017-09-15", "2017-10-17", 
"2018-02-22", "2017-12-04", "2017-01-18", "2017-09-15", "2018-05-13", 
"2017-09-15", "2017-03-07", "2018-05-13")), row.names = c(NA, 
-20L), class = c("tbl_df", "tbl", "data.frame"))

Answer 1

meant 的意思是每种美学都要使用一次。最好是对您的前数据和发布数据进行bind_rows，然后绘制合并的数据：

library(tidyverse)
df <- bind_rows(list(pre = preMax, post = postMax), .id = 'time') 

ggplot() +
  geom_line(data = df, aes(x=buffer, y=max, color=date, linetype= time)) +
  guides(color = guide_legend(ncol = 2))

​

​

作为将日期放在颜色美学上的另一种选择，您可以使用面处理来保持清晰(虽然这看起来有点奇怪，但提供的样本有限，它应该给您一个大致的想法)：

ggplot() +
  geom_line(data = df, aes(x=buffer, y=max, linetype= time)) +
  facet_wrap(facets = ~date)

​

​