问题
I have a polar plot that graphs hourly data over a year. I have managed to put in four annotation rectangles to denote season. I would like these rectangles to have a gradient fill from clear to the current colour. Here is my current graph:
I have tried to put in a gradient fill for the rectangles specifically, but this conflicts with the marker scale fill gradient. Ideally the graph would look like this:
Here is my code so far:
#how to generate a dataset with hourly readings over a year and a half.
library(lubridate)
NoOfHours <- as.numeric(ymd_hms("2019-6-1 00:00:00") - ymd_hms("2018-3-1 00:00:00"))*24
data1 <- as.data.frame(ymd_hms("2018-3-01 8:00:00") + hours(0:NoOfHours))
colnames(data1) <- 'date'
set.seed(10)
data1$level <- runif(nrow(data1), min = 0, max = 400)
library(readxl);library(lubridate); #loads the 'readxl' package.
#1.
Hours <- format(as.POSIXct(strptime(data1$date,"%Y-%m-%d %H:%M:%S",tz="")) ,format = "%H:%M:%S")
data1$hours <- Hours
Date <- format(as.POSIXct(strptime(data1$date,"%Y-%m-%d %H:%M:%S",tz="")) ,format = "%Y-%m-%d")
data1$date_date <- Date#output
month <- format(as.POSIXct(strptime(data1$date,"%Y-%m-%d %H:%M:%S",tz="")) ,format = "%m-%d")
data1$month<- month
#input the date here to select the start of the dataset, use the format: "yyyy-mm-dd". Then choose the end date by taking one full year of data. I.E. start = "2018-3-1", end = "2019-2-28"
start <- ceiling_date(ymd(data1$date_date[1]), "day", change_on_boundary = FALSE)
startdate <- as.Date(start) %m+% days(1)
enddate1 <- as.Date(startdate) %m+% years(1)
enddate<- as.Date(enddate1) %m-% days(1)
devicenumber <- "1"
Housename <- "level.tiff"
houseinfo <- c(devicenumber, Housename)
graphlimit <- 0 #need to define a limit for the graph
i<-200 #the initial lowest limit will always be 200
#this loop will now check for the highest levels of Radon and then graph a graphlimit that will encompass this maxima. This newly determined limit will allow different datasets to easily be automatically plotted with a range that is not too big or too small for the data.
if (max(data1$level) < (i+50)) {
graphlimit <- i
} else {
while (max(data1$level)>(i+50)) {
i<-i+200 }
if(max(data1$level) < (i+50)) {graphlimit <- i
}
}
library(openair)
yeardata <- selectByDate(data1, start = startdate, end = enddate, year = 2018:2019) #select for a defined set of years
library(ggplot2);library(extrafont)
graphlength <- graphlimit/(1350/1750)
innerlimit <- -(graphlength*(200/1750))
plotlimit <- graphlength+innerlimit #this sets the end limit of the outer plot ticks. This ratio was determined based on the largest dataset.
starttimedate <- ymd_hms(paste(startdate, "01:00:00"))
endtimedate <- ymd_hms(paste(enddate1, "01:00:00"))
#endtimedate2 <- ymd_hms(paste(floor_date(ymd(data1$date_date[1]), "year"), "01:00:00"))
NoOfhours <- as.numeric(ymd_hms(starttimedate) - ymd_hms("2018-01-01 00:00:00"))*24
NoOfHours <- (8760/12)*(month(startdate)-1)#as.numeric(ymd_hms(starttimedate) - ymd_hms(endtimedate2))*24 #need this to determine rotation. This will determine how many hours are between Jan 1-1 at 0:0:0 till the start of the dataset.
NoOfHoursall <- as.numeric(ymd_hms(endtimedate) - ymd_hms(starttimedate))*24
date_vals <- seq(from = ceiling_date(ymd(startdate), "month", change_on_boundary = FALSE), length.out = 12, by = "months")
finalcell <- length(yeardata$date)
plot <- ggplot(yeardata, aes(x=date, y=level, color = level)) +
annotate("rect", xmin = ((yeardata$date[1])), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))]), ymin = graphlimit, ymax = Inf, fill = "springgreen4", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))]), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))]), ymin = graphlimit, ymax = Inf, fill = "goldenrod2", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))]), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))]), ymin = graphlimit, ymax = Inf, fill = "orangered3", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))]), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))]), ymin = graphlimit, ymax = Inf, fill = "cornflowerblue", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))]), xmax = (yeardata$date[finalcell]), ymin = graphlimit, ymax = Inf, fill = "springgreen4", alpha = 0.15)+
geom_hline(yintercept = seq(0, graphlimit, by = 200), colour = "black", size = 0.75, alpha = 0.3)+
geom_hline(yintercept = seq(0, graphlimit, by = 50), colour = "black", size = 0.5, alpha = 0.1)+
annotate("segment",x = (yeardata$date[1]), xend = (yeardata$date[1]), y = 0, yend = graphlimit, colour = "black", size = 1, alpha = 0.5) +
#annotate("text",x = (max(yeardata$date)), y = innerlimit, colour = "black", size = 7, alpha = 1, label = devicenumber)+
scale_colour_gradientn(limits = c(0,1000), colours = c("grey","yellow","orangered1","red","red4","black"), values = c(0,0.1,0.2,0.5,0.8,1), breaks = c(0, 100, 200, 500, 800, 1000), oob = scales::squish, name = expression(atop("",atop(textstyle("Level"^2*"")))))+ #need oob = scales::squish to get values over 200 to be red.
geom_jitter(alpha = 0.2, size = 1) +
theme(text = element_text(family="Calibri"), axis.title=element_text(size=16,face="bold"), axis.text.x = element_blank(), axis.text.y = element_text(size = 12))+
labs(x = NULL, y = bquote('Level'))+
scale_y_continuous(breaks = seq(0, graphlimit, 200),
limits = c(innerlimit,plotlimit))+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[1])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[1])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[2])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[2])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[3])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[3])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[4])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[4])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[5])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[5])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[6])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[6])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[7])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[7])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[8])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[8])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[9])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[9])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[10])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[10])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[11])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[11])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[12])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[12])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "01-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "JAN", angle = -15)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "02-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "FEB", angle = -45)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "03-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "MAR", angle = -74)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "04-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "APR", angle = -104)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "05-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "MAY", angle = -133)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "06-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "JUN", angle = -163)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "07-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "JUL", angle = 165)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "08-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "AUG", angle = 135)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "09-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "SEP", angle = 105)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "10-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "OCT", angle = 75)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "11-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "NOV", angle = 45)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "12-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "DEC", angle = 15)
plot
plot <- plot + coord_polar(start = ((2*NoOfhours/NoOfHoursall)*pi))+ #scale_x_continuous(breaks = as.POSIXct.Date(ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd(date_vals[2])))]), origin))+ #need to have the number of radians to get my start position. If march 1st is the start date, then 60 days have past since Jan 1.
theme(legend.title = element_text(color = "black", size = 14, face = "bold"), panel.background = element_rect(fill = "white"), panel.grid = element_blank())
plot
Any help would be much appreciated.
Thanks
回答1:
Well, after much looking, I have managed a solution. I found this post: How can I apply a gradient fill to a geom_rect object in ggplot2?
From that, I modified the answer given to include what is seen in my code below. Taking a quote from @baptiste: "you have two options: i) discretise the rectangles along y and map the fill or alpha to that variable; ii) post-process the plot e.g. via gridSVG, which supports natively gradient fills."
So essentially, I created a function that mapped transparency values to n number of rectangles. To get this to work with the different colours I wanted, I had to create a separate dataframe for each season, then within the function map each season to its own set of discretized rectangles with their specific colour. Here is the dataframe and function code specifically.
spring <- data.frame(matrix(ncol = 0, nrow = 1))
spring$seasonstartdate <- ymd_hms((yeardata$date[1]))
spring$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))])
spring$colour <- "springgreen4"
summer <- data.frame(matrix(ncol = 0, nrow = 1))
summer$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))])
summer$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))])
summer$colour <- "goldenrod2"
fall <- data.frame(matrix(ncol = 0, nrow = 1))
fall$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))])
fall$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))])
fall$colour <- "orangered3"
winter <- data.frame(matrix(ncol = 0, nrow = 1))
winter$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))])
winter$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))])
winter$colour <- "orangered3"
spring1 <- data.frame(matrix(ncol = 0, nrow = 1))
spring1$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))])
spring1$seasonenddates <- ymd_hms(yeardata$date[finalcell])
spring1$colour <- "springgreen4"
ggplot_grad_rects <- function(n, ymin, ymax) {
y_steps <- seq(from = ymin, to = ymax, length.out = n + 1)
alpha_steps <- seq(from = 0, to = 0.2, length.out = n)
rect_grad <- data.frame(ymin = y_steps[-(n + 1)],
ymax = y_steps[-1],
alpha = alpha_steps)
rect_total <- merge(spring, rect_grad)
rect_total2 <- merge(summer, rect_grad)
rect_total3 <- merge(fall, rect_grad)
rect_total4 <- merge(winter, rect_grad)
rect_total5 <- merge(spring1, rect_grad)
ggplot(yeardata)+
geom_rect(data=rect_total,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="springgreen4") +
geom_rect(data=rect_total2,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="goldenrod2") +
geom_rect(data=rect_total3,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="orangered3") +
geom_rect(data=rect_total4,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="cornflowerblue") +
geom_rect(data=rect_total5,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="springgreen4") +
guides(alpha = FALSE)
}
It turned out will in the end. Here is a plot that was created.
Now here is the full code so you all can see the process.
library(lubridate)
NoOfHours <- as.numeric(ymd_hms("2019-6-1 00:00:00") - ymd_hms("2018-3-1 00:00:00"))*24
data1 <- as.data.frame(ymd_hms("2018-3-01 8:00:00") + hours(0:NoOfHours))
colnames(data1) <- 'date'
set.seed(10)
data1$level <- runif(nrow(data1), min = 0, max = 400)
library(readxl);library(lubridate); #loads the 'readxl' package.
#1.
Hours <- format(as.POSIXct(strptime(data1$date,"%Y-%m-%d %H:%M:%S",tz="")) ,format = "%H:%M:%S")
data1$hours <- Hours
Date <- format(as.POSIXct(strptime(data1$date,"%Y-%m-%d %H:%M:%S",tz="")) ,format = "%Y-%m-%d")
data1$date_date <- Date#output
month <- format(as.POSIXct(strptime(data1$date,"%Y-%m-%d %H:%M:%S",tz="")) ,format = "%m-%d")
data1$month<- month
#input the date here to select the start of the dataset, use the format: "yyyy-mm-dd". Then choose the end date by taking one full year of data. I.E. start = "2018-3-1", end = "2019-2-28"
start <- ceiling_date(ymd(data1$date_date[1]), "day", change_on_boundary = FALSE)
startdate <- as.Date(start) %m+% days(1)
enddate1 <- as.Date(startdate) %m+% years(1)
enddate<- as.Date(enddate1) %m-% days(1)
devicenumber <- "1"
Housename <- "level.tiff"
houseinfo <- c(devicenumber, Housename)
graphlimit <- 0 #need to define a limit for the graph
i<-200 #the initial lowest limit will always be 200
#this loop will now check for the highest levels of Radon and then graph a graphlimit that will encompass this maxima. This newly determined limit will allow different datasets to easily be automatically plotted with a range that is not too big or too small for the data.
if (max(data1$level) < (i+50)) {
graphlimit <- i
} else {
while (max(data1$level)>(i+50)) {
i<-i+200 }
if(max(data1$level) < (i+50)) {graphlimit <- i
}
}
library(openair)
yeardata <- selectByDate(data1, start = startdate, end = enddate, year = 2018:2019) #select for a defined set of years
library(ggplot2);library(extrafont)
graphlength <- graphlimit/(1350/1750)
innerlimit <- -(graphlength*(200/1750))
plotlimit <- graphlength+innerlimit #this sets the end limit of the outer plot ticks. This ratio was determined based on the largest dataset.
starttimedate <- ymd_hms(paste(startdate, "01:00:00"))
endtimedate <- ymd_hms(paste(enddate1, "01:00:00"))
#endtimedate2 <- ymd_hms(paste(floor_date(ymd(data1$date_date[1]), "year"), "01:00:00"))
NoOfhours <- as.numeric(ymd_hms(starttimedate) - ymd_hms("2018-01-01 00:00:00"))*24
NoOfHours <- (8760/12)*(month(startdate)-1)#as.numeric(ymd_hms(starttimedate) - ymd_hms(endtimedate2))*24 #need this to determine rotation. This will determine how many hours are between Jan 1-1 at 0:0:0 till the start of the dataset.
NoOfHoursall <- as.numeric(ymd_hms(endtimedate) - ymd_hms(starttimedate))*24
date_vals <- seq(from = ceiling_date(ymd(startdate), "month", change_on_boundary = FALSE), length.out = 12, by = "months")
finalcell <- length(yeardata$date)
#HERE IS THE SOLUTION
#I created a few dataframes to represent the seasons with their start and end times. From there I modified a previous solution to create a gradient geom_rect function.
spring <- data.frame(matrix(ncol = 0, nrow = 1))
spring$seasonstartdate <- ymd_hms((yeardata$date[1]))
spring$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))])
spring$colour <- "springgreen4"
summer <- data.frame(matrix(ncol = 0, nrow = 1))
summer$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))])
summer$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))])
summer$colour <- "goldenrod2"
fall <- data.frame(matrix(ncol = 0, nrow = 1))
fall$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))])
fall$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))])
fall$colour <- "orangered3"
winter <- data.frame(matrix(ncol = 0, nrow = 1))
winter$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))])
winter$seasonenddates <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))])
winter$colour <- "orangered3"
spring1 <- data.frame(matrix(ncol = 0, nrow = 1))
spring1$seasonstartdate <- ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))])
spring1$seasonenddates <- ymd_hms(yeardata$date[finalcell])
spring1$colour <- "springgreen4"
ggplot_grad_rects <- function(n, ymin, ymax) {
y_steps <- seq(from = ymin, to = ymax, length.out = n + 1)
alpha_steps <- seq(from = 0, to = 0.2, length.out = n)
rect_grad <- data.frame(ymin = y_steps[-(n + 1)],
ymax = y_steps[-1],
alpha = alpha_steps)
rect_total <- merge(spring, rect_grad)
rect_total2 <- merge(summer, rect_grad)
rect_total3 <- merge(fall, rect_grad)
rect_total4 <- merge(winter, rect_grad)
rect_total5 <- merge(spring1, rect_grad)
ggplot(yeardata)+
geom_rect(data=rect_total,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="springgreen4") +
geom_rect(data=rect_total2,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="goldenrod2") +
geom_rect(data=rect_total3,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="orangered3") +
geom_rect(data=rect_total4,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="cornflowerblue") +
geom_rect(data=rect_total5,
aes(xmin=ymd_hms(seasonstartdate), xmax=ymd_hms(seasonenddates),
ymin=ymin, ymax=ymax,
alpha=alpha), fill="springgreen4") +
guides(alpha = FALSE)
}
plot <- ggplot_grad_rects(100, graphlimit, graphlength) +
annotate("rect", xmin = ((yeardata$date[1])), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))]), ymin = graphlimit, ymax = Inf, fill = "springgreen4", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2018-6-1")))]), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))]), ymin = graphlimit, ymax = Inf, fill = "goldenrod2", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2018-9-1")))]), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))]), ymin = graphlimit, ymax = Inf, fill = "orangered3", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2018-12-1")))]), xmax = (yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))]), ymin = graphlimit, ymax = Inf, fill = "cornflowerblue", alpha = 0.15)+
annotate("rect", xmin = (yeardata$date[min(which(yeardata$date_date == ymd("2019-3-1")))]), xmax = (yeardata$date[finalcell]), ymin = graphlimit, ymax = Inf, fill = "springgreen4", alpha = 0.15)+
geom_hline(yintercept = seq(0, graphlimit, by = 200), colour = "black", size = 0.75, alpha = 0.3)+
geom_hline(yintercept = seq(0, graphlimit, by = 50), colour = "black", size = 0.5, alpha = 0.1)+
annotate("segment",x = (yeardata$date[1]), xend = (yeardata$date[1]), y = 0, yend = graphlimit, colour = "black", size = 1, alpha = 0.5) +
#annotate("text",x = (max(yeardata$date)), y = innerlimit, colour = "black", size = 7, alpha = 1, label = devicenumber)+
scale_colour_gradientn(limits = c(0,1000), colours = c("grey","yellow","orangered1","red","red4","black"), values = c(0,0.1,0.2,0.5,0.8,1), breaks = c(0, 100, 200, 500, 800, 1000), oob = scales::squish, name = expression(atop("",atop(textstyle("Level"^2*"")))))+ #need oob = scales::squish to get values over 200 to be red.
geom_jitter(alpha = 0.2, size = 1) +
theme(text = element_text(family="Calibri"), axis.title=element_text(size=16,face="bold"), axis.text.x = element_blank(), axis.text.y = element_text(size = 12))+
labs(x = NULL, y = bquote('Level'))+
scale_y_continuous(breaks = seq(0, graphlimit, 200),
limits = c(innerlimit,plotlimit))+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[1])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[1])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[2])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[2])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[3])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[3])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[4])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[4])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[5])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[5])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[6])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[6])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[7])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[7])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[8])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[8])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[9])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[9])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[10])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[10])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[11])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[11])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("segment", x = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[12])))]), xend = (yeardata$date[min(which(yeardata$date_date == ymd(date_vals[12])))]), y = graphlimit, yend = plotlimit, colour = "black", size = 2)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "01-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "JAN", angle = -15)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "02-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "FEB", angle = -45)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "03-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "MAR", angle = -74)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "04-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "APR", angle = -104)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "05-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "MAY", angle = -133)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "06-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "JUN", angle = -163)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "07-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "JUL", angle = 165)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "08-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "AUG", angle = 135)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "09-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "SEP", angle = 105)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "10-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "OCT", angle = 75)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "11-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "NOV", angle = 45)+
annotate("text", x = (yeardata$date[min(which(yeardata$month == "12-16"))]), y = ((graphlimit+plotlimit)/2), colour = "black", size = 9, family="Calibri", label = "DEC", angle = 15)
plot
plot <- plot + coord_polar(start = ((2*NoOfhours/NoOfHoursall)*pi))+ #scale_x_continuous(breaks = as.POSIXct.Date(ymd_hms(yeardata$date[min(which(yeardata$date_date == ymd(date_vals[2])))]), origin))+
theme(legend.title = element_text(color = "black", size = 14, face = "bold"), panel.background = element_rect(fill = "white"), panel.grid = element_blank())
plot
Thanks and enjoy
来源：https://stackoverflow.com/questions/58122423/how-to-gradient-fill-an-annotation-shape-in-ggplot2