Creating a line plot using plot_ly (R) with two y-axes

Question

I am a newcomer to the plot_ly package and am trying to produce a time series line plot with two variables on the y axis.

In my dataframe 'baro' I have 'DateTime' variable in POSIXct format, and 'Pressure' and 'Temperature' in numeric format.

I am basing my code off the example given here: https://plot.ly/r/multiple-axes/

p <- plot_ly(baro)

add_trace(p, x = ~DateTime, y = ~Pressure, type = "scatter",
          mode = "lines", name = "Pressure")

add_trace(p, x = ~DateTime, y = ~Temperature, type = "scatter",
          mode = "lines", name = "Temperature", yaxis = "y2")

layout(p,
  title = "Pressure & Temperature", yaxis2 = ay,
  xaxis = list(title="x")
)

This outputs a set of axes labelled -1 to 6 on the x axis and -1 to 4 on the y axis with no data plotted.

Answer 1

I prefer use pipes %>% rather than attribute an object to a plot. When you have 2 Y-axis it's nice to set the layout of every one explicitly.

This should do what you want:

# Build randon data
set.seed(123)

baro = data.frame(DateTime = as.POSIXct(1:10,origin = "2019-01-01"),
                  Pressure = sample(1000:2000,10),
                  Temperature = sample(20:60,10)
                  )

# Build plot

baro %>%
  plot_ly(type = "scatter", mode = "lines") %>%
  add_trace(x = ~DateTime, y = ~Pressure, name = "Pressure")%>%
  add_trace(x = ~DateTime, y = ~Temperature, name = "Temperature", yaxis = "y2") %>%
  layout(title = "Pressure & Temperature",
         yaxis = list(title = "Pressure"),
         yaxis2 = list(title = "Temperature",
                       overlaying = "y",
                       side = "right"
                       )
         )

Here the output:

Best regards.