Hello. I was trying to determine worker’s home cbg. As mentioned in Patterns dataset document, the worker& non-worker filter has been removed since only a limited number of workers can be determined. However, I notice the neighborhood sample dataset has a column work_behavior_device_home_areas. If I did not misunderstand, it seems that the dataset provides the home cbg information for workers in each cbg. Is neighborhood dataset the only way I can get workers’ home cbg info? I was also wondering how reliable the work_behavior_device_home_areas feature is, since the work& non-worker filter has been removed from the POI-Pattern dataset. Thanks for help.
This topic was automatically generated from Slack. You can find the original thread here.
Hey @Sisi_Zhou - that’s generally the right idea. Certainly will not be perfect. The idea behind work_behavior_device_home_areas is that if a device is dwelling for 6+ hours then it’s an appropriate proxy for a work location.
In the past, we did try to filter out employee visits in our Patterns product, but it introduced too much noise. Therefore, for our Patterns (monthly/weekly) product we generally recommend filtering out visits in the highest bucket in bucketed_dwell_times. Thanks for the question!