Question about the Overview files:
- num_unique_visitors in [visit_panel_summary.csv] is NOT the sum of POI num_unique_visitors for that same time period and geography right? In the former, a unique visitor to multiple POIs is not counted multiple times correct?
- Why aren’t number_devices_residing from [home_panel_summary.csv] equal to num_unique_visitors in [visit_panel_summary.csv]? One possible answer is that there are some people who stay home the whole month and do not add to the number of unique visitors. However, sometimes the number of unique visitors is larger than the number of devices ( 39,944,996 total devices and 41,599,633 in March 2019 for example).
- A) yes it is possible for a device to have a home location, but not to have any visits
B) It is also possible for a device to have visits but not be included in home_panel_summary. Devices are only counted in home_panel_summary if we are able to assign that device a home_census area based on their historical data, aggregating over many weeks.
There are 2 reasons why a device may have a visit in Patterns but not be assigned a home location
- That device is not from the US. So when we analyzed their data to assign a home location, we determined that home was outside the US.
- That device did not have enough data to calculate a home area (but did have enough data to measure a visit). This happens sometimes as devices come in and out of our panel.
This is also why for an individual record in Patterns the raw_visitor_counts does not always equal the sum of visitors in visitor_home_cbgs, because the latter only counts visitors with a US home cbg.
Hope that answers your questions @Luis_E_Quintero – let me know if not or if you have more questions
Thanks Ryan. As usual, you were very helpful!