X-axis for times - r

Question

I am trying to generate a series of plots that show the same patient taking drinks and urinating at different times. Each plot represents a single day. I want to compare the days and hence I need to ensure that all graphs plotted have the same x-axis. My code is below which I cribbed from How to specify the actual x axis values to plot as x axis ticks in R
### Data Input
time_Thurs <- c("01:10", "05:50", "06:00","06:15", "06:25", "09:35", "10:00", "12:40",
"14:00", "17:20", "18:50", "19:10", "20:10", "21:00", "22:05", "22:35")
event_Thurs <- c("u", "u", "T", "T", "u", "u", "T","T","u", "u", "T", "T", "T", "T", "u", "W")
volume_Thurs <- c(NA, NA, 0.25, 0.25, NA, NA, 0.125, 0.625, NA, NA, 0.25, 0.25, 0.25, 0.25,
NA, 0.25)
total_liquids_Thurs <- sum(volume_Thurs, na.rm=TRUE)
time_Thurs <- paste("04/04/2019", time_Thurs, sep=" ")
time_Fri <- c("01:15", "06:00", "06:10", "06:25", "06:30", "07:10", "08:40", "09:20",
"12:45", "13:45")
event_Fri <- c("u","u", "T","T","u","uu","T", "u", "T", "u")
volume_Fri <- c(NA, NA, 0.25, 0.25, NA, NA, 0.125, NA, 0.625, NA)
total_liquids_Fri <- sum(volume_Fri, na.rm=TRUE)
time_Fri <- paste("05/04/2019", time_Fri, sep=" ")
### Collect all data together
event <- c(event_Thurs, event_Fri)
Volume <- c(volume_Thurs, volume_Fri)
time_log <- c(time_Thurs, time_Fri)
time_log <- strptime(time_log, format = "%d/%m/%Y %H:%M")
time_view <- format(time_log, "%H:%M")
### Put into Dataframe
patient_data <- data.frame(time_log, time_view, event, Volume)
# write.csv(patient_data, file="patient_data.csv", row.names = FALSE)
daily_plot <- function(x, day) {
# x patient data - a data.frame with four columns:
# POSIXct time, time, event and Volume
# date number of day of month
# y volume of liquid
# TotVol total volume of intake over week
# Event - drink or otherwise
x <- x[as.numeric(format(x[,1], "%d")) == day, ]
TotVol <- sum(x[,4], na.rm = TRUE)
DayOfWeek <- weekdays(x[1,1], abbreviate = FALSE)
plot(x[,1],x[,4],
xlim = c(x[1,1],x[length(x[,1]),1]),
xlab="Hours of Study", ylab = "Volume of Liquid Drank /L",
main = paste("Total Liquids Drank = ", TotVol, " L on ", DayOfWeek, "Week 1, Apr 2019"),
sub = "dashed red line = urination", pch=16,
col = c("black", "yellow", "green", "blue")[as.numeric(x[,3])],
xaxt = 'n'
)
xAxis_hrs <- seq(as.POSIXct(x[1,1]), as.POSIXct(x[length(x[,1]),1]), by="hour")
axis(1, at = xAxis_hrs, las = 2)
abline( v = c(x[x[,3] == "u",1]), lty=3, col="red")
}
When I run the function,
daily_plot(patient_data, 4)
I want to print out my x-axis, as amended in the form of hours representing the events over the 24 hour period.
When I wrap my xAxis_hrs vector in strptime(xAxis_hrs, format = "%H") the code crashes - that is the x-axis doesn't print out and I see, Error in axis(1, at = xAxis_hrs, las = 2) : (list) object cannot be coerced to type 'double' . Any help?

Answer 1

The issue is that you pass the labels to the wrong named argument, namely at (which should be the numeric positions of the labels). Use the following instead:
axis(1, at = xAxis_hrs, labels = strptime(xAxis_hrs, format = "%H"), las = 2)
Unfortunately this doesn’t change the fact that the axis labels don’t fit into the plot, and collide with the axis title. The former can be fixed by adjusting the plot margins. I’m not aware of a good solution for the latter, although changing the time format might help: it’s probably not necessary/helpful to print the full minutes and seconds (which are always 0). In fact, did you mean to use format instead of strptime?
Apart from that I fundamentally agree with the other answer recommending ggplot2 in the long run. It makes this kind of stuff a lot less painful.

Answer 2

If you're open to a ggplot solution:
library(tidyverse)
library(lubridate)
daily_ggplot <- function(df, selected_day) {
df_day <- filter(df, day(time_log) == selected_day)
df_urine <- filter(df_day, event == "u")
df_drink <- filter(df_day, event != "u")
TotVol <- sum(df_day$Volume, na.rm = TRUE)
Date <- floor_date(df_day$time_log[1], 'days')
DayOfWeek <- weekdays(Date, abbreviate = F)
plot_title <- paste0("Total drank = ", TotVol, "L on ", DayOfWeek, " Week 1, Apr 2018")
ggplot(df_drink) +
aes(time_log, Volume, color = event) +
geom_point(size = 2) +
geom_vline(data = df_urine, aes(xintercept = time_log), color = "red", linetype = 3) +
labs(x = "Hours of Study", ylab = "Volume of Liquid Drank (L)",
title = plot_title, subtitle = "lines = urination") +
theme_bw() +
scale_x_datetime(date_labels = "%H:%M", limits = c(Date, Date + days(1)))
}
daily_ggplot(patient_data, 4)