I have data set that track daily activity of user. User repeat action one or more times during a day. This actions are independent and I want to determine if there is relation between times of events.

day one First event: 08.14
Second event: 09:00
Third event: 11.45
Fourth event: 12:00

day two First event: 08.05
Second event: 09:50
Third event: 11.05

and so on... as you can see number of actions during a day is determine by distribution that is known. Since I know distribution how I can model this times of events.

******************* adding plots of data ********************

enter image description here

enter image description here

  • $\begingroup$ So you want to check whether a user repeats these actions at roughly the same time of the day using data from many days? What is this known distribution that you speak of? $\endgroup$ – user2974951 Jan 11 at 9:02
  • $\begingroup$ Negative binomial distributon $\endgroup$ – explorer Jan 11 at 9:09
  • $\begingroup$ Can you post two images of your data, 1) a histogram of all the event times for all the days combined, and 2) a frequency plot of daily events, that is group all the days into one and in each day count the number of events and draw a histogram / barplot. $\endgroup$ – user2974951 Jan 11 at 13:12
  • $\begingroup$ @user2974951 I hope that this is what you were asking, this are data for all days. First chart show that there are 72 days with just one action that day.Second chart I choose bin width arround 5 minutes. $\endgroup$ – explorer Jan 11 at 13:41

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.