Closed srt1 closed 7 years ago
I'm confused. The schema has a definition for a variable that indicates seasonal adjustment in the data (https://lehd.ces.census.gov/data/schema/V4.2a-draft/lehd_public_use_schema.html#_seasonadj) and for the file name (https://lehd.ces.census.gov/data/schema/V4.2a-draft/lehd_csv_naming.html#_sa). The schema makes no statement (because it is a schema...) about exactly which data series have or have not seasonal adjustment.
-- Lars Vilhuber, Economist Cornell University, Executive Director, Labor Dynamics Institute and ILR School - Department of Economics
e: lars.vilhuber@cornell.edu p: +1.607-330-5743 v: https://cornell.zoom.us/my/larsvilhuber w: http://lars.vilhuber.com/ http://lars.vilhuber.com/
Assistant: ldi@cornell.edu | +1.607-255-2744
GnuPG Fingerprint: 0D7D 527F 9268 F693 74BB A666 FD01 37F0 3362 7346
From: srt1 notifications@github.com Sent: Wednesday, August 16, 2017 8:39:51 AM To: labordynamicsinstitute/qwi_schemas Cc: Lars Vilhuber; Mention Subject: Re: [labordynamicsinstitute/qwi_schemas] Add something about which series are seasonally adjustment to the schema? (#64)
Assigned #64https://github.com/labordynamicsinstitute/qwi_schemas/issues/64 to @larsvilhuberhttps://github.com/larsvilhuber.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/labordynamicsinstitute/qwi_schemas/issues/64#event-1208191883, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AGsoeMd2OsIxW2-Ryc2c4GQJq11H4LxRks5sYuMXgaJpZM4O42cd.
I suppose one thing that the schema does not describe is exactly what set of tables are provided in the release. Users can reference the naming scheme to see if a table is there, and whether that table is seasonally adjusted or not. The label_agg_level.csv table does flag whether the aggregation exists on the release. It does not indicate whether it is available in seasonally adjusted form. If it's not necessary to do so, we don't have to include it. Hence the question mark in the issue title. Users can be left to figure that out.
... and hence my confusion on #23. If we are not detailing exactly which tables are being presented in the schema, I don't see a reason to do so in the Excel files; and since the Excel files are otherwise just the csv files themselves, plus a few lines of headers, I don't see very much there to actually document. The header lines themselves are (for example):
Source: United States Census Bureau Release: R2017Q2 Data Schema: V4.2b-draft Tabulation: by state, all firms, all workers (Not Seasonally Adjusted)
If you like, feel free to set up a template of what you are looking for there, and I can fill in the missing info.
@srt1 @heathhayward Re: documenting seasonal adjustment. Both the contents of a data series (the variable seasonadj
) and the naming convention (the suffix sa
, if present in the naming convention) should be sufficient for this purpose.
The schema presently says (https://lehd.ces.census.gov/data/schema/V4.2a-draft/lehd_csv_naming.html#_basic_schema) that J2J in all its variants has a suffix in the name indicating whether or not a particular file contains ONLY seasonally adjusted, or ONLY non-adjusted time series. I believe that is sufficient.
@srt1 Please close if you are in agreement.
Closing per discussion, not explicitly including the set of tables/margins available with seasonally adjusted counterparts as part of the schema at this time.
The schema at this time does not indicate which tables/series have seasonally adjusted complements, and which do not. The current rule is as follows: