How to get the smooth line for monthly rainfall using ggplot?

Question

I am trying to plot the monthly rainfall data from 1986 to 2016 using ggplot. My dataframe looks like this:

head(df)
  Year Month Station Rainfall   Remarks
1 1986   Jan    stn1      0.0  Observed
2 1986   Feb    stn1     10.4  Observed
3 1986   Mar    stn1     16.5  Estimated
4 1986   Apr    stn1     34.0  Observed
5 1986   May    stn1     27.0  Observed
6 1986   Jun    stn1    159.4  Observed

str(df)
'data.frame':   1488 obs. of  5 variables:
 $ Year    : chr  "1986" "1986" "1986" "1986" ...
 $ Month   : Ord.factor w/ 12 levels "Jan"<"Feb"<"Mar"<..: 1 2 3 4 5 6 7 8 9 10 ...
 $ Station : Factor w/ 4 levels "stn1","stn2",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ Rainfall: num  0 10.4 16.5 34 27 ...
 $ Remarks : Factor w/ 2 levels "Estimated","Observed": 2 2 1 2 2 2 2 2 2 2 ...

I tried the following code:

library(ggplot2)
ggplot(df, aes(x=Year, y=Rainfall, col=Station)) + geom_line()

However the above code results in vertical lines plot, while I want to have smooth varying lines.

I want to plot all the four station (stn1 to stn4) such that the color of each line be based on the df$Remarks. Also is it possible to have unique color for each station?

Your help would be appreciated

Answer 1

Here is one approach if you create a month-year variable:

library(ggplot2)
library(zoo)

df$Mo_Yr <- as.yearmon(paste0(df$Year, '-', df$Month), "%Y-%b")

ggplot(df, aes(x=Mo_Yr, y=Rainfall, col=Station)) + 
  geom_line() +
  scale_x_yearmon()

If you want to use different color points for Remarks (Observed and Estimated), for a single Station, you could try the following:

ggplot(df, aes(x=Mo_Yr, y=Rainfall)) +
  geom_point(aes(col = Remarks)) +
  geom_line() +
  scale_x_yearmon()

If you want to plot 2 lines for Observed and Estimated, you could add col argument to geom_line as below. Note I added some example data to illustrate. Depending on what data you have available this may (or may not) be what you need.

ggplot(df, aes(x=Mo_Yr, y=Rainfall)) +
  geom_line(aes(col=Remarks)) +
  scale_x_yearmon()

Data (for last example)

df <- read.table(text = 
"Year Month Station Rainfall   Remarks
 1986   Jan    stn1      0.0  Observed
 1986   Feb    stn1     10.4  Observed
 1986   Mar    stn1     16.5  Estimated
 1986   Apr    stn1     34.0  Observed
 1986   May    stn1     27.0  Observed
 1986   Jun    stn1    159.4  Observed
 1986   Jul    stn1     83.1  Estimated
 1986   Aug    stn1     55.7  Observed
 1986   Sep    stn1     12.3  Estimated", header = T, stringsAsFactors = T)

Answer 2

You might want to try passing the stat_smooth parameter

ggplot(df) + 
  geom_line(aes(y= Rainfall, x= Year, color= Station)) +
  stat_smooth(aes(y= Rainfall, x= Year), method = lm, formula = y ~ poly(x, 10), se = FALSE)