How to display activity over time when data contains "activity start time" and "activity duration"

[pdumas]

Hello,

I am working on the help desk statistics.

The "Call" file contains :
Call_ID eg : 1
Agent_ID eg : Paul
Call_start_timestamp eg : 2009-12-15 10:25:35
Call_duration eg : 120 (seconds)

I have loaded a "Timeline" table with year/month/day/hour/min/sec fields calculated from "Call_Start_timestamp"

A chart with

Dimension : A cyclic group with year/month/...

Expression : count(distinct Agent_ID)

counts the Agents that began a call during a defined period (year/month/...)

What would be the best solution to count the Agents that where "On Call" during a defined period?

Thanks
Pierre

[prieper · Accepted solution]

Can you post some sample data or some scrambled data? Guess that you should be able to solve this problem with INTERVALMATCH-function.

Peter

[pdumas]

Please find a sample qvw

With Intervalmatch function, I suppose I need to generate a "OnCall" table with 1 line par second for each agent "on call" and use a chart expression like "count(distinct agent_id)" to find the number of agents simultaneously on call.

This will generate sevral million lines per month and I may face performance problems.

Thanks for your help.

Pierre.

[prieper]

Do you really need to break this down into seconds?

[pdumas]

Minutes should sufficient. hours are not from a user perspective.

However,if you have a solution, even partial, I will manage to improve it or negociate with users.

Thanks for your help

Pierre.

[#6]

Hi Pierre

Did you get an answer on this one? i have the very same problem and would be glad to get an answer on it.

I want to count the numer of Check-in Counter(Schalter) that are used simutaneously over 30 Minutes...

I tried something like:

if(LoginTime <= [30 Min] and( LogoffTime >= [30 Min]), count(Distinct Schalter),0) //where [30 Min] is my interval field.

or

=count({<LoginTime *= {"<= $ [30 Min]"}, LogoffTime *= {">=$ [30 Min]"}>} Distinct Schalter)

but both solutions won't work...

Thanks for any helpful reply.

Kind regards,

Anthony