I need to determine probability when user create first file during a day. I have time stamp in format HH:MM:SS. of creation first file during one year period. In order to determine distribution I created histogram in a way to create binds (10 minutes range) and count frequencies. Based on histogram this could be a gamma distribution as possible solution. For gamma I need mean and standard deviation.

Question is can I convert HH:mm:ss time stamp to seconds and calculate mean and standard deviation.

Similar research I found on internet about baby delivery but no methodology explained.

enter image description here

this graph show 8 h of working day activities ( 08:00 - 16:00) , based on 10 minutes range - 10 hours - if some one came earlier or stay longer and produce file.

  • $\begingroup$ Could you clarify if you want to determine the most likely time when a user will create their first new file of the day or the probability distribution for when a user will create their first new file of the day? Also, do you want to do this for single user or for a population of users? $\endgroup$ Jan 4, 2018 at 1:52
  • $\begingroup$ I want to determine the most likely time when will user create first file of the day. Number of files during a day follow negative binomial distribution. So I will use pseudo random generator based on this distribution for determine how many files will user create. Second step is to determine when will this happened durring a day. $\endgroup$
    – explorer
    Jan 4, 2018 at 7:46
  • $\begingroup$ @Machineepsilon I want to model times of creation for group of users (same type of user - regular user ). When I use pseudo random generator to generate arrow of 100 numbers that follow certain distribution (in my case negative binomial distribution for number of files for 100 days of that user, example: 1,2,1,4,2,5,.... 2) every time it will produce another sequence. Now i need to add timestamp when first file will be generate. I need to produce additional arrow with times so I can say arrow of files[1] combine with arrow of times[1] and so on.... I hope that I manage to explain $\endgroup$
    – explorer
    Jan 4, 2018 at 10:30
  • $\begingroup$ @Tim, can I apply solution from this post in my case? $\endgroup$
    – explorer
    Jan 4, 2018 at 11:04


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.