Plotting data against time in R

Question

I have a data frame with one column as date/time(stored internally as numeric) and other columns as numeric/integer and I want to plot data against date/time.

Date/time in the data frame was populated using the following.

as.POSIXct(strptime(time, '%H:%M:%S %p %m/%d/%Y',tz='GMT')) 

class(table$time) is numeric.

1. How do I plot and show the data in x axis as readable date time in some format.
2. How do I plot subset of rows and not all rows Ex: rows between dateTime1 and dateTime2 where dateTime1 and dateTime2 are date given in a certain format.

Answer 1

Here is some dummy data:

data <- structure(list(time = structure(c(1338361200, 1338390000, 1338445800, 1338476400, 1338532200, 1338562800, 1338618600, 1338647400, 1338791400, 1338822000), class = c("POSIXct", "POSIXt"), tzone = ""), variable = c(168L, 193L, 193L, 201L, 206L, 200L, 218L, 205L, 211L, 230L)), .Names = c("time", "variable"), row.names = c(NA, -10L), class = "data.frame")
data
              time variable
1  2012-05-30 09:00:00      168
2  2012-05-30 17:00:00      193
3  2012-05-31 08:30:00      193
4  2012-05-31 17:00:00      201
5  2012-06-01 08:30:00      206
6  2012-06-01 17:00:00      200
7  2012-06-02 08:30:00      218
8  2012-06-02 16:30:00      205
9  2012-06-04 08:30:00      211
10 2012-06-04 17:00:00      230

To show date and time in the axis you can use function axis.POSIXct:

plot(data, xaxt="n")
axis.POSIXct(side=1, at=cut(data$time, "days"), format="%m/%d") 

You can control where the ticks fall with at (as for regular functionaxis except here it is to be provided with objects of class POSIXct) and control how they will appear with format.

As far as subsetting is concerned, as long as your dateTime1 and dateTime2 objects are also POSIXct objects you can do it as you would do any other kind of subsetting.

dateTime1 <- strptime("00:00 05/31/2012", format="%H:%M %m/%d/%Y")
dateTime2 <- strptime("3 Jun 2012 05-30", format="%d %b %Y %H-%M")
data[data$time < dateTime2 & data$time > dateTime1, ]
                 time variable
3 2012-05-31 08:30:00      193
4 2012-05-31 17:00:00      201
5 2012-06-01 08:30:00      206
6 2012-06-01 17:00:00      200
7 2012-06-02 08:30:00      218
8 2012-06-02 16:30:00      205

Answer 2

You could also use ggplot2, more specifically geom_point or geom_line (note that I use the example data from @plannapus):

require(ggplot2)
theme_set(theme_bw()) # Change the theme to my preference
ggplot(aes(x = time, y = variable), data = data) + geom_point()

or using a line geometry:

ggplot(aes(x = time, y = variable), data = data) + geom_line()

ggplot2 automatically recognizes that the datatype of the x-axis is a date, and draws the axis accordingly.