r - Histogram over time of day -

i have dataframe in 1 column contains different dates including times, e.g.

as.posixct(c("2017-07-03 08:23:00",              "2017-07-03 09:00:00",              "2017-07-03 17:23:00",              "2017-07-03 18:05:00",              "2017-07-04 08:24:00",              "2017-07-04 09:02:00",              "2017-07-04 17:24:00",              "2017-07-04 18:01:00",              "2017-07-05 08:57:00",              "2017-07-05 09:31:00",              "2017-07-05 16:25:00",              "2017-07-05 17:14:00")) 

now want @ how many times time occurs @ intervals (say 15 min). thus, aim histogram of frequency (over days) vs time of day.

any hints?

edit: tried extract time by

df$time <- hm(format(df$date, "%h:%m")) 

but left me column of class period didn't know how handle. tried like

ggplot(df, aes(date)) + geom_histogram() + scale_x_time() 

my main problem here how can use ggplot plotting.

probably easier way this, approach...

library(plyr) library(lubridate)  #sample data df<-data.frame(time=as.posixct(c(   '2017-07-03 08:23:00',   '2017-07-03 09:00:00',   '2017-07-03 17:23:00',   '2017-07-03 18:05:00',   '2017-07-04 08:24:00',   '2017-07-04 09:02:00',   '2017-07-04 17:24:00',   '2017-07-04 18:01:00',   '2017-07-05 08:57:00',   '2017-07-05 09:31:00',   '2017-07-05 16:25:00',   '2017-07-05 17:14:00')))  #extract time df$hour = hour(df$time) + minute(df$time)/60 + second(df$time)/3600  #create bins bins=c(paste0(rep(c(paste0(0,0:9),10:23), each=4),".", c("00",25,50,75))[-1],"24:00")  #divide data bins df$bins = cut(df$hour, breaks=seq(0, 24, 0.25), labels=bins)  #reformat numeric df$bins <- as.numeric(as.character(df$bins))  #histogram hist(df$bins)  #with ggplot library(ggplot2) ggplot(df, aes(bins)) +   geom_histogram()