R,使用scale_linetype_manual更改ggplot图例名称 [英] R, change ggplot legend names with scale_linetype_manual
问题描述
我有一个看起来像这样的数据框:
>df年份mpft值类型索引1 1996 2 0.033827219固体2.12 1997 2 0.133278701固体2.13 1998 2 0.261428650固体2.14 1999 2 0.394702438固体2.15 1996 3 0.019079686固体3.16 1997 3 0.074332942固体3.17 1998 3 0.149042964固体3.18 1999 3 0.227812452固体3.19 1996 4 0.009909126实心4.110 1997 4 0.026231721固体4.111 1998 4 0.052912805固体4.112 1999 4 0.086256016固体4.113 1996 17 0.017256492固体17.114 1997 17 0.079446280固体17.115 1998 17 0.166014538实心17.116 1999 17 0.316175339固体17.117 1996 18 0.080072523固体18.118 1997 18 0.313289644实心18.119 1998 18 0.629398957实心18.120 1999 18 1.024946245实心18.1110 1996 2 0.031634282虚线2.221 1997 2 0.139244701虚线2.231 1998 2 0.273270126破折号2.241 1999 2 0.412409808虚线2.251 1996 3 0.019430502破折号3.261 1997 3 0.079252516破折号3.271 1998 3 0.161607337破折号3.281 1999 3 0.252595611破折号3.291 1996 4 0.009976637破4.2101 1997 4 0.027057403破折号4.2111 1998 4 0.055755671破折号4.2121 1999 4 0.093064641破4.2171 1996 18 0.061041422破折号18.2181 1997 18 0.245554619破折号18.2191 1998 18 0.490633135破折号18.2201 1999 18 0.758070060破折号18.2
我正在尝试绘制数据并具有正确的图例,到目前为止,我最初已经尝试过
ggplot(df,aes(x = Year,y = value,color = factor(mpft),linetype = type))+geom_line(aes(group = index),size = 1.4)+#scale_linetype_manual(name ="Run Type",值=唯一(df $ type),labels = run.type)+scale_color_manual(name ="PFT",值= setNames(mycol [unique(df $ mpft)],唯一(df $ mpft)),标签= setNames(mynam [唯一(df $ mpft)],唯一(df $ mpft)))
哪个给我
我尝试用添加一个scale_linetype_manual
ggplot(df,aes(x = Year,y = value,color = factor(mpft),linetype = type))+geom_line(aes(group = index),size = 1.4)+scale_linetype_manual(name ="Run Type",值= unique(df $ type),labels = run.type)+scale_color_manual(name ="PFT",值= setNames(mycol [unique(df $ mpft)],唯一(df $ mpft)),标签= setNames(mynam [唯一(df $ mpft)],唯一(df $ mpft)))
使用
>运行类型[1]当前""origED3"
但是我得到
,它具有图例的正确名称,但具有不同的线型.我想念什么?
编辑
我数据框的 dput
是
>dput(df)结构(列表(年= c(1996,1997,1998,1999,1996,1997,1998,1999、1996、1997、1998、1999、1996、1997、1998、1999、1996、1997,1998、1999、1996、1997、1998、1999、1996、1997、1998、1999、1996,1997、1998、1999、1996、1997、1998、1999),mpft = c(2L,2L,2L,2L,3L,3L,3L,3L,4L,4L,4L,4L,17L,17L,17L,17L,18L,18L,18L,18L,2L,2L,2L,2L,3L,3L,3L,3L,4L,4L,4L,4L,18L,18L,18L,18L),值= c(0.0338272191848643,0.133278701149992,0.261428650232716、0.394702437670559、0.0190796862689925、0.0743329421068756,0.149042964352043、0.227812451937011,0.00990912614900737、0.0262317206863519,0.0529128049802722、0.0862560162908444、0.017256491619149、0.0794462797803606,0.166014537897384、0.31617533869767、0.0800725232220131、0.31328964372358,0.629398957462415、1.02494624459608、0.0316342818911836、0.139244700529005,0.273270126484303、0.412409807917143、0.0194305022713642、0.0792525159706922,0.161607337403947、0.252595610607411,0.00997663742883768、0.0270574028188436,0.0557556714277292、0.0930646413413941、0.0610414215913856、0.245554619318541,0.490633135315979,0.758070059865948),类型=结构(c(1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,1L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L),.Label = c("solid","dashed"),class ="factor"),索引=结构(c(1L,1L,1L,1L,2L,2L,2L,2L,3L,3L,3L,3L,4L,4L,4L,4L,5L,5L,5L,5L,6L,6L,6L,6L,7L,7L,7L,7L,8L,8L,8L,8L,9L,9L,9L,9L),. Label = c("2.1","3.1","4.1","17.1","18.1","2.2","3.2","4.2","18.2"),class ="factor")),.Names = c("Year","mpft","value","type","index"),row.names = c("1","2","3","4","5","6","7","8","9","10","11","12","13","14","15","16","17","18","19","20","110","21","31","41","51","61","71","81","91","101","111","121","171","181","191","201"),类别="data.frame")
编辑
这不是一个非常优雅的解决方案,而是将破折号"
替换为"22"
,然后使用:
ggplot(df,aes(x = Year,y = value,color = factor(mpft),linetype = type))+geom_line(aes(group = index),size = 1.4)+scale_linetype_manual(name ="Run Type",值=唯一(as.character(df $ type)),labels = run.type)+scale_color_manual(name ="PFT",值= setNames(mycol [unique(df $ mpft)],唯一(df $ mpft)),标签= setNames(mynam [唯一(df $ mpft)],唯一(df $ mpft)))
我能够做到正确
有趣的问题,也是我以前从未见过的一个问题.
简短答案
差异是因为您的第一个版本中的线型根本不是虚线"
.
好答案
在您的第一个版本中,没有为 linetype
美感指定任何内容,ggplot默认为 scale_linetype_discrete()
和
I have a dataframe which looks like this:
> df
Year mpft value type index
1 1996 2 0.033827219 solid 2.1
2 1997 2 0.133278701 solid 2.1
3 1998 2 0.261428650 solid 2.1
4 1999 2 0.394702438 solid 2.1
5 1996 3 0.019079686 solid 3.1
6 1997 3 0.074332942 solid 3.1
7 1998 3 0.149042964 solid 3.1
8 1999 3 0.227812452 solid 3.1
9 1996 4 0.009909126 solid 4.1
10 1997 4 0.026231721 solid 4.1
11 1998 4 0.052912805 solid 4.1
12 1999 4 0.086256016 solid 4.1
13 1996 17 0.017256492 solid 17.1
14 1997 17 0.079446280 solid 17.1
15 1998 17 0.166014538 solid 17.1
16 1999 17 0.316175339 solid 17.1
17 1996 18 0.080072523 solid 18.1
18 1997 18 0.313289644 solid 18.1
19 1998 18 0.629398957 solid 18.1
20 1999 18 1.024946245 solid 18.1
110 1996 2 0.031634282 dashed 2.2
21 1997 2 0.139244701 dashed 2.2
31 1998 2 0.273270126 dashed 2.2
41 1999 2 0.412409808 dashed 2.2
51 1996 3 0.019430502 dashed 3.2
61 1997 3 0.079252516 dashed 3.2
71 1998 3 0.161607337 dashed 3.2
81 1999 3 0.252595611 dashed 3.2
91 1996 4 0.009976637 dashed 4.2
101 1997 4 0.027057403 dashed 4.2
111 1998 4 0.055755671 dashed 4.2
121 1999 4 0.093064641 dashed 4.2
171 1996 18 0.061041422 dashed 18.2
181 1997 18 0.245554619 dashed 18.2
191 1998 18 0.490633135 dashed 18.2
201 1999 18 0.758070060 dashed 18.2
I am trying to plot the data and have the right legend, so far I have initially tried with
ggplot(df,aes(x=Year,y=value, colour = factor(mpft),linetype=type)) +
geom_line(aes(group = index), size = 1.4) +
#scale_linetype_manual(name= "Run Type", values = unique(df$type), labels = run.type) +
scale_color_manual(name = "PFT",
values = setNames(mycol[unique(df$mpft)], unique(df$mpft)),
labels = setNames(mynam[unique(df$mpft)], unique(df$mpft)))
Which gives me
I have tried adding a scale_linetype_manual with
ggplot(df,aes(x=Year,y=value, colour = factor(mpft),linetype=type)) +
geom_line(aes(group = index), size = 1.4) +
scale_linetype_manual(name= "Run Type", values = unique(df$type), labels = run.type) +
scale_color_manual(name = "PFT",
values = setNames(mycol[unique(df$mpft)], unique(df$mpft)),
labels = setNames(mynam[unique(df$mpft)], unique(df$mpft)))
with
> run.type
[1] "current" "origED3"
But I get
which has the right names for the legend but has a different linetype. What am I missing?
EDIT
The dput
of my dataframe is
> dput(df)
structure(list(Year = c(1996, 1997, 1998, 1999, 1996, 1997, 1998,
1999, 1996, 1997, 1998, 1999, 1996, 1997, 1998, 1999, 1996, 1997,
1998, 1999, 1996, 1997, 1998, 1999, 1996, 1997, 1998, 1999, 1996,
1997, 1998, 1999, 1996, 1997, 1998, 1999), mpft = c(2L, 2L, 2L,
2L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L, 17L, 17L, 17L, 17L, 18L,
18L, 18L, 18L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 4L, 4L, 4L, 4L,
18L, 18L, 18L, 18L), value = c(0.0338272191848643, 0.133278701149992,
0.261428650232716, 0.394702437670559, 0.0190796862689925, 0.0743329421068756,
0.149042964352043, 0.227812451937011, 0.00990912614900737, 0.0262317206863519,
0.0529128049802722, 0.0862560162908444, 0.017256491619149, 0.0794462797803606,
0.166014537897384, 0.31617533869767, 0.0800725232220131, 0.31328964372358,
0.629398957462415, 1.02494624459608, 0.0316342818911836, 0.139244700529005,
0.273270126484303, 0.412409807917143, 0.0194305022713642, 0.0792525159706922,
0.161607337403947, 0.252595610607411, 0.00997663742883768, 0.0270574028188436,
0.0557556714277292, 0.0930646413413941, 0.0610414215913856, 0.245554619318541,
0.490633135315979, 0.758070059865948), type = structure(c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L,
2L, 2L, 2L), .Label = c("solid", "dashed"), class = "factor"),
index = structure(c(1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 3L, 3L,
3L, 3L, 4L, 4L, 4L, 4L, 5L, 5L, 5L, 5L, 6L, 6L, 6L, 6L, 7L,
7L, 7L, 7L, 8L, 8L, 8L, 8L, 9L, 9L, 9L, 9L), .Label = c("2.1",
"3.1", "4.1", "17.1", "18.1", "2.2", "3.2", "4.2", "18.2"
), class = "factor")), .Names = c("Year", "mpft", "value",
"type", "index"), row.names = c("1", "2", "3", "4", "5", "6",
"7", "8", "9", "10", "11", "12", "13", "14", "15", "16", "17",
"18", "19", "20", "110", "21", "31", "41", "51", "61", "71",
"81", "91", "101", "111", "121", "171", "181", "191", "201"), class = "data.frame")
EDIT
Not a very elegant solution but substituting "dashed"
with "22"
and then using:
ggplot(df,aes(x=Year,y=value, colour = factor(mpft),linetype=type)) +
geom_line(aes(group = index), size = 1.4) +
scale_linetype_manual(name= "Run Type", values = unique(as.character(df$type)), labels = run.type) +
scale_color_manual(name = "PFT",
values = setNames(mycol[unique(df$mpft)], unique(df$mpft)),
labels = setNames(mynam[unique(df$mpft)], unique(df$mpft)))
I am able to get it right
Interesting question, and one I haven't come across on SO before.
Short answer
The difference is because the linetype in your first version isn't "dashed"
at all.
Long answer
In your first version, without specifying anything for the linetype
aesthetic, ggplot defaults to scale_linetype_discrete()
, and the current code for that is:
scale_linetype <- function(..., na.value = "blank") {
discrete_scale("linetype", "linetype_d", linetype_pal(),
na.value = na.value, ...)
}
Hence, scale_linetype_discrete()
gets its linetype values from the linetype_pal
function, courtesy of the scales
package (at least, that's the only place I found it):
> scales::linetype_pal()(2)
[1] "solid" "22"
When you specified the linetype
aesthetic mapping in the second version, using scale_linetype_manual()
, the corresponding current code is:
scale_linetype_manual <- function(..., values) {
manual_scale("linetype", values, ...)
}
Thus when you explicitly ask for c("solid", "dashed")
as the two linetype values in your plot, ggplot uses them. When you don't, the default values are c("solid", "22")
, and "22"
corresponds to a different, more tightly spaced pattern than "dashed"
's pattern.
Demonstration below, using built-in data:
df.sample <- diamonds %>%
filter(cut %in% c("Fair", "Good")) %>%
group_by(cut, clarity) %>%
summarise(price = mean(price / carat)) %>%
ungroup()
p <- ggplot(df.sample,
aes(x = clarity, y = price, group = cut,
linetype = cut)) +
geom_line(size = 1) +
guides(linetype = guide_legend(keywidth = 3, keyheight = 1)) +
theme(legend.position = c(1, 0), legend.justification = c(1, 0))
library(gridExtra)
grid.arrange(p +
labs(title = "Default scale",
subtitle = c("values = linetype_pal()(2)")),
p + scale_linetype_manual(values = c("solid", "dashed")) +
labs(title = "Manual scale",
subtitle = "values = c('solid', 'dashed')"),
p + scale_linetype_manual(values = c("solid", "22")) +
labs(title = "Manual scale",
subtitle = "values = c('solid', '22')"),
nrow = 1)
The third plot mimics the behaviour of the default scale.
这篇关于R,使用scale_linetype_manual更改ggplot图例名称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IT屋!