Multiple y axis for bar plot and line graph using ggplot

Question

I have some transpiration data from an experiment that I would like to show as a time series on a line graph using R. I also have some precipitation data which I would like

Answer 1

Multiple axis are not available in ggplot, it's a choice of Hadley Wickham, the creator of the package. You just can't.

Source: https://stackoverflow.com/a/3101876/8031980

Answer 2

As described with nice examples here or here starting with version 2.2.0 of ggplot2, is possible to add a secondary axis.

One could try this:

Prepare the provided data for ggplot:

# Read data from OP's DropBox links
am_means <- read.csv("https://www.dropbox.com/s/z4dl0jfslhqccb8/am_means.csv?dl=1")
rainfall <- read.csv("https://www.dropbox.com/s/vkv9vm5o93ttk1i/Rainfall.csv?dl=1")

am_means$dates <- as.Date(am_means$dates, format = "%d/%m/%Y")
rainfall$DATE <- as.Date(rainfall$DATE,format = "%d/%m/%Y")

# join the two tables
my_data_all <- merge(x = am_means,
                     y = rainfall,
                     by.x = "dates",
                     by.y = "DATE",
                     all = TRUE)

# use data between desired date interval (rainfall had some extra dates)
require(data.table)
setDT(my_data_all)
my_data <- my_data_all[dates %between% c("2017-01-31", "2017-04-06")]

Plot with secondary OY axis:

Is important to transform the data appearing on 2nd OY (right-hand side). Since the max value is approx 2 times smaller that the one of the data from first OY axis (left-hand side), one can multiply by 2.

my_factor <- 2
my_plot <- ggplot(my_data, 
                  aes(x = dates,
                      group = Treatment)) +
    geom_errorbar(aes(ymin = cond - err,
                      ymax = cond + err,
                      colour = Treatment),
                  width = 0.5,
                  size  = 0.5) +
    geom_line(aes(y = cond, colour = Treatment)) + 
    geom_point(aes(y = cond, colour = Treatment)) +
    # here the factor is applied 
    geom_bar(aes(y = daily.rainfall * my_factor), 
             fill = "light blue",
             stat = "identity")
my_plot

Adding the 2nd OY axis with scale_y_continuous. Note the transformation back. If above we multiplied by 2, now we divide data by 2:

my_plot <- my_plot + scale_y_continuous(sec.axis = sec_axis(trans = ~ . / my_factor, 
                                                            name = "Rainfall (mm)"))

Continues with OP's code

my_plot <- my_plot + scale_x_date(date_breaks = "2 weeks",
                                  date_minor_breaks = "1 week",
                                  date_labels = "%b %d") + 
    scale_color_manual(breaks = c("Control", "T1", "T2", "T3", "T4"),
                       values = c("blue", "yellow", "hotpink1", "orange", "red")) +
    labs(title = "Stomatal Conductance - Tamata Maples", 
         y = expression(Conductance (m~mol~m^{-2})), 
         x = "Date") +
    theme_bw()
my_plot

Some other related answered questions: adding a cumulative curve or combining Bar and Line chart.