Closed MojcaFranic closed 5 years ago
I suggest we still decide on weather keep or drop null users. That will probably lead us to a different number again (as well as decision on what we will be dropping for our own analysis)
I would agree on dropping the test users.
Regarding null users, I'm not sure. I think, if you are concerned with click data and don't match it with the user, it does not matter if the user is null. But if you are concerned with the actual user data, it might be difficult.
I think we should keep the same pool of users for all visualisations so the hosts can visit our website and be able to clearly understand what each chat is saying and the context etc.
Side note: When I do user.shape I only get 22 users back. What am I doing wrong?
I noticed that some locations have values that suggest the user being only a test user and not accurate data so I will be dropping those as well