3

In the following plot I wanted to remove the date and keep the time in hours. I have seen few solutions like here and another where Time is converted to numeric. I liked the use of require(scales)

ggplot(data= dfweek2, aes(x = Time, y = Mean_steps)) + facet_grid(day ~.) + geom_path()

this is my plot enter image description here

When I change to require(scales) ggplot(data= dfweek2, aes(x = Time, y = Mean_steps)) + facet_grid(day ~.) + geom_path() + scale_x_datetime(labels = date_format("%H:%M"))

this scale is not longer correct...enter image description here

What am I doing wrong?

> str(dfweek2)
Classes ‘grouped_df’, ‘tbl_df’, ‘tbl’ and 'data.frame': 576 obs. of  4 variables:
 $ day       : Factor w/ 2 levels "Weekday","Weekend": 1 1 1 1 1 1 1 1 1 1 ...
 $ interval  : int  0 5 10 15 20 25 30 35 40 45 ...
 $ Mean_steps: num  2.333 0.462 0.179 0.205 0.103 ...
 $ Time      : POSIXct, format: "2015-06-15 00:00:00" "2015-06-15 00:05:00" ...
Community
  • 1
  • 1
pdm
  • 57
  • 5

1 Answers1

3

I wonder if it's a timezone thing. Look at the output of attributes(dfweek2$Time) and see if it has a tzone attribute. If it does, then when format is called without a tz parameter, it will convert the times from whatever time zone they're stored as to whatever system time zone you've got set. If this is the case, you have two options.

  1. Remove the tzone attribute from dfweek2$Time

    attr(dfweek2$Time, "tzone") <- NULL

  2. Provide the formatter with the tz parameter. For example, if you find the time zone of dfweek2$Time is "EST", then:

    scale_x_datetime(labels = function(x) format(x, "%H:%M", tz = "EST"))

Nick Kennedy
  • 12,510
  • 2
  • 30
  • 52
  • thanks Nick. Both are insightful solutions. I also found that I had not set `tz` parameter, so it was taking local tz and and the `date_format` was taking `GMT` as default. Hence, the issue. – pdm Jun 15 '15 at 16:12