2 Replies Latest reply on Jul 28, 2016 5:23 AM by Mohamed Malik

    Need help filtering outliers from rest of dataset in my graph

    Mohamed Malik

      Hello everyone,


      I am having difficulties filtering a set of outlying data from my graph.


      I am currently looking at a circle graph and on my graph there are 6 clear outliars. I want to remove them from the rest of the dataset.


      Here is a link to the exercise I am following




      It is Exercise 1, specifically the part regarding Sign In Count / DogID.


      The pill is known as "Sign In Count" since it is the number of instanches users have signed-in to their account.


      DogID is reference to the ID of the dog related to each sign-in activity session.


      I'm basically meant to group and filter the outliers from the rest of the datasets, after applying Dog ID as a detail.


      When I highlight the 6 circles of outlying data (those with a sign-in count of more than 175), I right click and "Group" them. I then take the group and drag them to the Filter card, thus creating a filter.

      Afterwards, I have two fields:

      1. The Dog ID (Group)
      2. Other

      When I turn off the 'Other' filter, I am left with a graph that looks purely at the Dog ID (Group) - the 6 outliers. If I turn off the Dog ID Group filter and turn on the 'Other' filter, then I am looking at all the data - including the 6 outliers that were formally grouped.

      Why is this? I thought by creating a group for the outliers, it would exclude their respective data from everything else. But all it did was provide me a filter that allows me to focus on the outliers, as oppose to creating a filter where I can see all of the data NOT including the outliers. What am I suppose to do?


      Here is a link to an album that shows what I see:


      Imgur: The most awesome images on the Internet