Plotting multiple time series plots using median data in R

Question

Im just beginning to learn R so this might be redundant so I apologize. I am wanting to make this attach excel graph in R. My data includes daily water level reading for a reservoir from 1951 to 2016. I need the data broken into three series (1951 to 2013, 2014 to 2015, and 2016). I would like to plot the median values for these series per calendar day. And I would like to remove febuary 29 from the median values. Here is a link to my data:https://docs.google.com/spreadsheets/d/1u1Whfp6VHXkZgrC0sVn_mT9XiVxszMhqlAszjZXzM1E/edit?usp=sharing

Here is what I have so far:

LL <- read.csv("BSLL.csv")
str(LL)
LLpre <- filter(LL, year > "1952" & year <"2014") 
headtail(LLpre, n=3)
medianLLpre = ddply(LLpre, .(month, day), summarise, level = median(level),   na.rm = FALSE)
LLpost <- filter(LL, year > "2013" & year < "2016")
headtail(LLpost, n=3)
medianLLpost = ddply(LLpost, .(month, day), summarise, level = median(level), na.rm = FALSE)
LL2016 <- filter(LL, year == "2016")
headtail(LL2016, n=3)
medianLL2016 = ddply(LL2016, .(month, day), summarise, level = median(level), na.rm = FALSE)

Excel Plot

JustGettinStarted JustGettinStarted · Accepted Answer · 2017-01-24T16:37:49

library(data.table)
Data <- fread("BSLL - BSLL.csv")

Create temporary time interval categories 1=1951 to 2013,2=2014 to 2015, and 3=2016

Data[,Categories:=ifelse(as.numeric(year)<2014,1,
                         ifelse(as.numeric(year)<2016,2,3))]

Create temporary data with no feb 29th

Temp.Data <- Data[!month==2][!day==29]

Compute the median by categories and plot.

Medians <- Temp.Data[,median(as.numeric(level)),by=.(month,Categories)]
plot(x=Medians[,month], y=Medians[,V1],type="n", xlab="", ylab="", 
     xaxt="n", yaxt="n", bty="l")
axis(2, tick=TRUE, labels=FALSE)
axis(2, tick=FALSE, labels=TRUE, line=-0.5)
axis(1, tick=TRUE, labels=FALSE)
axis(1, tick=FALSE, labels=TRUE, line=-0.5)
lines(x=Medians[Categories==1][,month],
      y=Medians[Categories==1][,V1],type="l",lwd=1,col="red")
lines(x=Medians[Categories==2][,month],
      y=Medians[Categories==2][,V1],type="l",lwd=1,col="green")
lines(x=Medians[Categories==3][,month],
      y=Medians[Categories==3][,V1],type="l",lwd=1,col="blue")
legend('topright','groups',
       legend=c("1951 to 2013","2014 to 2015","2016"), bty="n",lty=c(1,1),
       col=c("red","green","blue"))

How's this? You can tinker with the details of the plot (sizes, positions, colours, labels, etc.. yourself)

P.S in future its not a good idea to release your full data set publicly. Contributors here just need an example of your data to work with and test

Plotting multiple time series plots using median data in R

1 Answers