R：如何只在x轴上打印时间而忽略日期

Question

I have an xts object, a snapshot of the data is as follow:

ts <- structure(c(620519.432512971, 619086.596917204, 620478.784694651, 
620997.044378227, 620885.262877848, 620275.545603053), index = structure(c(1519954199.9845, 
1519955999.7115, 1519957799.9675, 1519959599.9935, 1519961399.9365, 
1519963199.2225), tzone = "", tclass = c("POSIXct", "POSIXt")), .indexCLASS = c("POSIXct", 
"POSIXt"), .indexTZ = "", tclass = c("POSIXct", "POSIXt"), tzone = "", class = c("xts", 
"zoo"), .Dim = c(6L, 1L), .Dimnames = list(NULL, "yy"))

All the timestamps are within the same date , and I am trying to add more tick marks on the x axis (time), which I managed to achieve. It looks as follow:

ggplot(data=ts) + 
geom_line(aes(x=Index,y=yy,colour = "yy"),na.rm=T) + 
scale_x_datetime(date_breaks ="20 min")

As you can see the timestamps above are correct, but i want to remove the date part and just have the time on the x-axis. So I tried the following code:

ggplot(data=ts) + 
geom_line(aes(x=Index,y=yy,colour = "yy"),na.rm=T) + 
scale_x_datetime(date_breaks ="20 min",labels = date_format("%H:%M:%S"))

But the times are all incorrect. What should I do to fix this to make it print only the time, leaving out the date, since the xts data are all within the same date ?

Answer 1

I believe your times are being converted to UTC times when you plot them. You can convert to your timezone. Here is how I convert to using the "US/Pacific" time zone. I added the tz argument to date_format

ggplot(data=ts) + 
     geom_line(aes(x=Index,y=yy,colour = "yy"),na.rm=T) + 
     scale_x_datetime(date_breaks ="20 min",
                      labels = date_format("%H:%M:%S", tz = "US/Pacific"))

You may have to change the timezone (see OlsonNames()

Answer 2

It is very simple,

ggplot(data=ts) + 
        geom_line(aes(x = Index,y = yy,colour = "yy"), na.rm=T) + 
        scale_x_datetime(date_breaks = "20 min", date_labels = "%H:%M:%S")

You just have to use parameter date_labels to set the formatting option of your datetime object.

Sources:
- http://ggplot2.tidyverse.org/reference/scale_date.html
- https://www.stat.berkeley.edu/~s133/dates.html