34 Replies Latest reply: Sep 15, 2016 9:47 PM by Sunny Talwar RSS

    Upper and lower bound limit

    Fikri Shihab

      Hi Qlikview experts,

       

      I have set of data several days worth of data with millions of records that have time stamps (e.g. every minute of 5 minute intervals) and several columns containing various information.

       

      To simplify, the columns of interest are:

       

      Time Stamp, User Name, network id, user status, region, throughput

       

      Network status value is either IN or OUT

       

      I am trying to create a chart showing number of distinct users accessing the network over the time (e.g. on hourly basis) or with variable interval and plotting upper and lower bound limit (w/ 95% confidence  interval). In this case user status will need to be IN.

       

      What I did was

       

      Dimension = Hour (Time Stamp)

       

      Expression is

       

      count({<[user_status]={'*IN*'}>}distinct([User Name]))

       

      While  I can simply show the above expression in the chart, the lower and upper limit w/95% confidence is not properly shown.

       

      Any suggestion how this should be done. Is there any aggregation function that should to be used here?

       

      If I wanted to do the same thing for throughput (e.g. using sum(throughput)), will it be the same expression to be used?

      Showing

       

      Also I'd like to be able to show the same thing when changing the dimension to a calculated dimension like below

       

      Timestamp((Ceil([Time Stamp],1/IntervalSize)),'DD-MMM-hh:mm')

       

      The interval size for various period (15 min, 30 min , 1 hour, etc.) is predefined in the loading script.