Forum Discussion
Looking at call data using Poisson Distribution, is their a better way
Hi
I am making my first steps into more statistical/forecast reporting using power bi. Currently I am looking at the companies telephone system and how many calls we take in an attempt to help with handling call volumes.
Looking at the data over a date range of a few months I realised that the report needed to allow for peak and off peak times.
So I am looking at each hour of the day indvidually.
I created 24 measures that are the mean average for each hour of the day.
then I will use POISSON.DIST() function the first parameter is the number of occurances I want to predict the probability for.
I intend to cover from 40 - 180 in groups of 5.
So I will need to create a measure for each of these groups and then again for each hour of the day.
Is their a better way to create this ? I fell calaculation groups could help me but I am unsure how I would set this up.
Can a more efficent way be set up?
You shouldn't need the 24 separate measures for each hour. The simplest way would be to make a Time table that has a column for hours, half hours, and any other time intervals you need. You can then use those columns in your visual. If you provide some sample data in a copy/paste-able format (or a link to it), a more specific solution can be provided.
Pat
1 Reply
You shouldn't need the 24 separate measures for each hour. The simplest way would be to make a Time table that has a column for hours, half hours, and any other time intervals you need. You can then use those columns in your visual. If you provide some sample data in a copy/paste-able format (or a link to it), a more specific solution can be provided.
Pat