在ggplot2中绘制矩形 - 无效的输入：time_trans仅适用于类POSIXct的对象 [英] Plotting rectangles in ggplot2 - Invalid input: time_trans works with objects of class POSIXct only
问题描述
df< - structure（list（IndID = structure（c（16L，15L，14L ，13L，12L，11L，
10L，9L，8L，7L，6L，5L，4L，3L，2L，1L）。标签= c（16，15，
14，13，12，11，10，9，8，7，6，5，4，3 StartDate = structure（c（1313042400，
1312956000，1313560800，1363672800，1374040800,1374040800,1374040800，
1374040800,1374040800,1374040800 ，1374040800,1365832800,1365919200，
1366178400,1395727200,1395813600），class = c（POSIXct，POSIXt
）），EndDate = structure（c（1377928800,1378015200,1378015200，
1386572400,1410760800,1410760800,1410760800,1410674400,1410760800，
1406959200,1399356000,1427868000,1394517600,1428213600,1428040800，
1420959600），class = c（POSIXct，POSIXt）） ），.Names = c（IndID，
StartDate，EndDate），row.names = c（NA，-16L），class =data.frame）
IndID StartDate EndDate
1 1 2011-08- 11 2013-08-31
2 2 2011-08-10 2013-09-01
3 3 2011-08-17 2013-09-01
4 4 2013-03-19 2013 -12-09
5 5 2013-07-17 2014-09-15
6 6 2013-07-17 2014-09-15
并且可以做这个阴谋
library（lubridate）
require（gglopt2）
df $ IndID < - factor（df $ IndID，levels = rev（df $ IndID））
p1< - ggplot（df，aes（x = IndID，fill = IndID））+
geom_rect（aes（x = IndID，xmin = as.numeric（IndID） - 0.45，xmax = as.numeric（IndID）+ 0.45 ，ymin = StartDate，ymax = EndDate））+
coord_flip（）+
xlab（IndID）+
ylab（Time）+
ggtitle期间）+
geom_text（aes（y = StartDate + as.difftime（8，unit =weeks），label = paste（month（StartDate，label = T），year（StartDate），sep = - ）））+
geom_text（aes（y = EndDate - as.difftime（9，unit =weeks），label = paste（month（EndDate，label = T），year（EndDate） ，sep = - ）））
p1
另外，我想在每年的6月19日至10月19日之间遮蔽该地区。为此，我创建了一个 data.frame 日期，然后在新的 data.frame 。 （是的，这是笨重的..）
temp< - data.frame（
start = as.Date （c（'2011-06-19'，'2012-06-19'，'2013-06-19'，'2014-06-19'，'2015-06-19'）），
end = as.Date（c（'2011-10-19'，'2012-10-19'，'2013-10-19'，'2014-10-19'，'2015-10-19'）））
str（temp）
dateRanges< - data.frame（
start = as.POSIXct（temp [，1]，％Y-％m-％d）+ hours（6 ），
end = as.POSIXct（temp [，2]，％Y-％m-％d）+ hours（6））
str（dateRanges）
当我尝试使用下面的代码将新的矩形添加到图中时，我收到了帖子标题中指出的错误。 p>
p1 + geom_rect（data = dateRanges，aes（xmin = start，xmax = end，ymin = -Inf，ymax = Inf）， inherit.aes = F，alpha = 0.4，fill = c（lightblue））
到目前为止正如我可以通过查看dateRanges的str（）所了解的，它们被正确地格式化为POSIXct类。
我做了请参阅
编辑：为了让背后的酒吧需要一点点摆弄。首先，我们必须调用geom_rect，它使得酒吧落后，并且我们将xmin和xmax从aes调用中移出，以避免与轴和因素级别相混淆。
p1 < - ggplot（df，aes（x = IndID，fill = IndID））+
geom_rect（data = dateRanges，aes（ymin = start，ymax = end），xmin = -Inf，xmax = Inf，alpha = 0.4，inherit.aes = FALSE，fill =lightblue）+
geom_rect（data = df，aes（x = IndID，xmin = as。数字（IndID） - 0.45，xmax = as.numeric（IndID）+ 0.45，ymin = StartDate，ymax = EndDate））+
coord_flip（）+
xlab（IndID）+
ylab（Time）+
ggtitle（Individual Monitoring Periods）+
geom_text（aes（y = StartDate + as.difftime（8，unit =weeks），label = paste month（StartDate，label = T），year（StartDate），sep = - ）））+
geom_text（aes（y = EndDate - as.difftime（9，unit =weeks），label =粘贴（month（EndDate，label = T），year（EndDate），sep = - ）））
p1
I have this data
df <- structure(list(IndID = structure(c(16L, 15L, 14L, 13L, 12L, 11L,
10L, 9L, 8L, 7L, 6L, 5L, 4L, 3L, 2L, 1L), .Label = c("16", "15",
"14", "13", "12", "11", "10", "9", "8", "7", "6", "5", "4", "3",
"2", "1"), class = "factor"), StartDate = structure(c(1313042400,
1312956000, 1313560800, 1363672800, 1374040800, 1374040800, 1374040800,
1374040800, 1374040800, 1374040800, 1374040800, 1365832800, 1365919200,
1366178400, 1395727200, 1395813600), class = c("POSIXct", "POSIXt"
)), EndDate = structure(c(1377928800, 1378015200, 1378015200,
1386572400, 1410760800, 1410760800, 1410760800, 1410674400, 1410760800,
1406959200, 1399356000, 1427868000, 1394517600, 1428213600, 1428040800,
1420959600), class = c("POSIXct", "POSIXt"))), .Names = c("IndID",
"StartDate", "EndDate"), row.names = c(NA, -16L), class = "data.frame")
IndID StartDate EndDate
1 1 2011-08-11 2013-08-31
2 2 2011-08-10 2013-09-01
3 3 2011-08-17 2013-09-01
4 4 2013-03-19 2013-12-09
5 5 2013-07-17 2014-09-15
6 6 2013-07-17 2014-09-15
and can make this plot
library(lubridate)
require(gglopt2)
df$IndID <- factor(df$IndID, levels = rev(df$IndID))
p1 <- ggplot(df, aes(x=IndID, fill = IndID))+
geom_rect(aes(x = IndID, xmin = as.numeric(IndID) - 0.45, xmax = as.numeric(IndID) + 0.45, ymin = StartDate, ymax = EndDate))+
coord_flip()+
xlab("IndID")+
ylab("Time")+
ggtitle("Individual Monitoring Periods")+
geom_text(aes(y = StartDate + as.difftime(8, unit = "weeks"), label = paste(month(StartDate, label = T), year(StartDate), sep = "-"))) +
geom_text(aes(y = EndDate - as.difftime(9, unit = "weeks"), label = paste(month(EndDate, label = T), year(EndDate), sep = "-")))
p1
I addition, I want to shade the region between June 19 and October 19 for every year. To do so, I made a data.frame of dates which were then converted to POSIXct format in a new data.frame. (yes, this is clunky..)
temp <- data.frame(
start = as.Date(c('2011-06-19', '2012-06-19', '2013-06-19', '2014-06-19', '2015-06-19')),
end = as.Date(c('2011-10-19', '2012-10-19', '2013-10-19', '2014-10-19', '2015-10-19')))
str(temp)
dateRanges <- data.frame(
start = as.POSIXct(temp [,1], "%Y-%m-%d") + hours(6),
end = as.POSIXct(temp [,2], "%Y-%m-%d") + hours(6))
str(dateRanges)
When I try to add the new rectangles to the plot with the following code, i get the error indicated in the post title.
p1 + geom_rect(data = dateRanges, aes(xmin = start , xmax = end, ymin = -Inf, ymax = Inf), inherit.aes= F, alpha = 0.4, fill = c("lightblue"))
As far as I can tell by looking at the str() of the dateRanges, they are correctly formatted as POSIXct class.
I did see the similar post here but am still struggling to connect the dots between the to related issues and fix my problem. Thanks in advance for any suggestions.
Your problem is you ran coord_flip() in your initial call. The plot is flipped, but ggplot still thinks that the x and y are the original x and y.
So, to fix, just switch your x and y aes in the final geom_rect:
p1 + geom_rect(data = dateRanges, aes(ymin = start , ymax = end, xmin = -Inf, xmax = Inf), inherit.aes= F, alpha = 0.4, fill = c("lightblue"))
p1
EDIT: To make the bars behind requires a little bit of fiddling. first we have to call the geom_rect which makes the bars so it is behind, and we move the xmin and xmax out of the aes call to avoid messing with the axis and levels of factors:
p1 <- ggplot(df, aes(x=IndID, fill = IndID))+
geom_rect(data = dateRanges, aes(ymin = start, ymax = end), xmin = -Inf, xmax = Inf, alpha = 0.4, inherit.aes=FALSE, fill = "lightblue")+
geom_rect(data = df, aes(x = IndID, xmin = as.numeric(IndID) - 0.45, xmax = as.numeric(IndID) + 0.45, ymin = StartDate, ymax = EndDate))+
coord_flip()+
xlab("IndID")+
ylab("Time")+
ggtitle("Individual Monitoring Periods")+
geom_text(aes(y = StartDate + as.difftime(8, unit = "weeks"), label = paste(month(StartDate, label = T), year(StartDate), sep = "-"))) +
geom_text(aes(y = EndDate - as.difftime(9, unit = "weeks"), label = paste(month(EndDate, label = T), year(EndDate), sep = "-")))
p1
这篇关于在ggplot2中绘制矩形 - 无效的输入：time_trans仅适用于类POSIXct的对象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！
