Reply To: Understanding structure for time series

#591
Andee Rubin
Participant
4 pts

I wanted to add an observation to this thread.  A central part of the issue is that there are really THREE variables here: year, state and rate – notice that in any of your data tables in the original dataset, any row has a mixture of values.  In one table, both the state name and the rates are in a row (and the years are the columns); in another table, both the year and the rates are in a row (and the states are the columns).   There’s something about the structure of the dataset that requires a something other than a flat structure.  The attribute stacker fixes the problem, but it would help to have a more semantic description of the problem that it fixes and the way it does so.  We wouldn’t necessarily share this reasoning with the middle school kids, but it’s very helpful for the staff of Data Clubs (and maybe also for CODAP folks).