Does anyone have any tips or useful functions for working with Neighborhood Patterns data in R? For instance, even trying to get stops_by_each_hour separated out at the date-time level vertically I’m struggling (same issue even if was trying to get stops_by_day). Something like:
Area Date time Visits
421010108001…4/1/2020 00:00:00 … 1
421010108001…4/1/2020 01:00:00 … 3
421010108001…4/1/2020 02:00:00 … 1
421010108001…4/1/2020 03:00:00 … 0
@Nick_H-K_Seattle_University would this be accomplished by the SafeGraphR package’s
expand_integer_json is the way to go.
Dang, I checked this like 3 times and convinced myself it didn’t apply. When I tried that, I get the error “Error: Argument ‘txt’ must be a JSON string, URL or file.” I’m sure a simple syntax error.
expand_integer_json(april2020, 'stops_by_each_hour', by = 'area'
Hmm, that’s odd, the syntax looks fine. Are there maybe weird additional quotes around the entries of stops_by_each_hour? Or non-numeric entries?
Stops_by_day also throws the same error. Thanks for all the help!
Hmm, that’s strange. I’ll take a look tonight
If useful, in trying to isolate/troubleshoot, it seems the “fromJSON” function used in your function is not reading the input as JSON for some reason. Even feeding it “fromJSON(april2020$stops_by_day)” I get the same error as above.
Converting the column to as.character() seemed to fix it! Thanks!
The next question I have is if areas are non-unique (e.g., because we have two months of data), what is the best way to tag the “date_range_start” or some other way to keep that indexed as well?
date_range_start, you can just add that to the
by option of
expand_integer_json , i.e.
by = c('area','date_range_start') and it will give you a separate row for each area for each date. You may want to convert
date_range_start from a datetime to a date with
as.Date() first since the hour information isn’t particularly informative.