[DDI-ADG] standards for frequency (or lack thereof)
Sandra.A.Cannon at frb.gov
Sandra.A.Cannon at frb.gov
Wed Aug 3 16:25:37 EDT 2005
Kate (and company),
Right now what lives in SDMX is a smaller set of what I listed. I've been
lobbying Arofan to expand things. For example, right now all observations
are dated as of the beginning of the period with full mm-dd-yyyy
specification (or maybe its yyyy-mm-dd, I can't remember) which means that
an observation which represents the second quarter of 2005 has a frequency
of quarterly but the date associated with it is actually April 1, 2005 -
the first day of the period. In general, this doesn't matter since we
translate any observation with a quarterly frequency to be represented by
the quarter in which the specific date falls.
Where it does matter for us is in the specification of weekly. Right now
there is only a weekly frequency which means that any data point specified
as being a weekly observation will (must!) be dated as representing the
Monday of a week. This is a HUGE problem for us at the Board because a
large number of our weekly interest rates are actually Wednesday
observations. Therefore, having a frequency of weekly isn't sufficient
since the requirement is that the date associated with an observation be at
a particular point in the period.
So to make a short story long, SDMX is looking to expand its frequency
coverage which will alleviate most (but probably not all) issues like this
but it is still important to remember that frequency does not equal time.
/san/
(who is coming off her soap box to pack for her visit to the in-laws then a
week of sun and sangria in Spain - but, alas, no TAB!)
-------------------------------------------------
San Cannon
Chief, Economic Information Management
Federal Reserve Board
Washington DC 20551
(202) 452-3710
scannon at frb.gov
Katherine
McNeill-Harman
<mcneillh at MIT.EDU To
> Mary Vardigan
<maryv at icpsr.umich.edu>,
08/03/2005 03:46 Sandra.A.Cannon at frb.gov
PM cc
ddi-adg at icpsr.umich.edu
Subject
Re: [DDI-ADG] standards for
frequency (or lack thereof)
And if these are used in SDMX, I imagine that they're machine actionable?
I'm wondering how this would work with my hope that someone could design a
search interface where the user could select among several pre-defined
frequency options; I was originally imagining a pull-down menu where the
user would be able to choose among annual, monthly, etc. But it sound like
this would maybe be more compatible with a search box where they could
format their search accordingly (e.g. put in Y10 for decennial).
I guess it's the difference between giving a set of discrete options vs.
defining a system of frequency naming conventions that could hopefully be
searched in a standard way. The one San has presented certainly provides a
lot more detail.
Kate
At 03:21 PM 8/3/2005 -0400, Mary Vardigan wrote:
>That sounds right. I was just trying to distinguish between an observation
>every 10 years and 10 years of observations. But if Y10 means the former,
>that's great. Thanks for the clarification.
>
>Mary
>
>At 02:54 PM 8/3/2005, Sandra.A.Cannon at frb.gov wrote:
>
>
>
>
>>Aren't they the same thing? If there is 10 years per period, doesn't
that
>>mean that there is an observation every 10 years?
>>
>>Here's how our documentation describes it:
>>
>>YPP(n, year, month)
>>Specifies data reported at intervals that are multiples of twelve months,
>>where n is a positive integer from 1 to 30, inclusive, indicating the
>>number of years per period. Year is a year number indicating a sample
>>ending year. Month is optional and is a month name indicating a sample
>>ending month. If month is omitted, DECEMBER is assumed.
>>
>>So if I ask for quarterly GDP to be displayed at the frequency
y(10,2000),
>>for example, I would get:
>>
>>1960 427.14
>>1970 768.55
>>1980 1839.50
>>1990 4367.40
>>2000 7742.50
>>
>>Isn't that what we are trying to do?
>>
>>I know that this is very time series oriented so let me know if I'm
missing
>>something important.
>>
>>/san/
>>
>>
>>-------------------------------------------------
>>San Cannon
>>Chief, Economic Information Management
>>Federal Reserve Board
>>Washington DC 20551
>>(202) 452-3710
>>scannon at frb.gov
>>
>>
>>
>>
>> Mary Vardigan
>> <maryv at icpsr.umic
>> h.edu>
To
>> Sent by: Sandra.A.Cannon at frb.gov
>> ddi-adg-bounces at i
cc
>> cpsr.umich.edu ddi-adg at icpsr.umich.edu
>>
Subject
>> Re: [DDI-ADG] standards for
>> 08/03/2005 02:44 frequency (or lack thereof)
>> PM
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>I wasn't sure if Y10 would indicate ten years of data since the
definition
>>says "YPP n years per period" or data collected every 10th year. But if
>>it's the latter, that would be great.
>>
>>Mary
>>
>>
>>At 02:39 PM 8/3/2005, Sandra.A.Cannon at frb.gov wrote:
>>
>>
>>
>>
>> >Mary,
>> >In the time series world, decennial wouldn't buy you much but in the
panel
>> >study world the same doesn't hold. Is a particular indicator for
>>decennial
>> >sufficient or would we be better to use the Yn indicator for such
things?
>> >Then Y10 would be decennial and Y2 would be biennial, etc. I
understand
>>if
>> >decennial is considered special enough to have its own code though.
>> >
>> >/san/
>> >
>> >
>> >-------------------------------------------------
>> >San Cannon
>> >Chief, Economic Information Management
>> >Federal Reserve Board
>> >Washington DC 20551
>> >(202) 452-3710
>> >scannon at frb.gov
>> >
>> >
>> >
>> >
>> > Mary Vardigan
>> > <maryv at icpsr.umic
>> > h.edu>
>>To
>> > Sandra.A.Cannon at frb.gov,
Katherine
>> > 08/03/2005 02:34 McNeill-Harman
<mcneillh at mit.edu>
>> > PM
>>cc
>> > ddi-adg at icpsr.umich.edu
>> >
>>Subject
>> > Re: [DDI-ADG] standards for
>> > frequency (or lack thereof)
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> >San, Kate, and others,
>> >
>> >This seems like a really good start on a controlled vocabulary for
>> >frequency. Do we need an additional indicator for something like the
>> >decennial U.S. Census?
>> >
>> >Mary
>> >
>> >At 04:24 PM 8/2/2005, Sandra.A.Cannon at frb.gov wrote:
>> >
>> >
>> >
>> >
>> > >Kate,
>> > >This may be where using common definitions from time series data
would
>>be
>> > >useful. There are many possible frequencies in the software that we
use
>> > >(which Arofan says they are contemplating building into SDMX). I
don't
>> > >know if it's authoritative or definitive but it certainly seems to be
>> > >exhaustive. I've attached a listing in a PDF of an internal webpage
I
>> > >support here.
>> > >
>> > >More food for thought....
>> > >
>> > >/san/
>> > >who is sorry she is going on holiday this week and will miss the next
>> > >several calls on this topic... well, not very sorry... ;-)
>> > >
>> > >(See attached file: freq.pdf)
>> > >
>> > >-------------------------------------------------
>> > >San Cannon
>> > >Chief, Economic Information Management
>> > >Federal Reserve Board
>> > >Washington DC 20551
>> > >(202) 452-3710
>> > >scannon at frb.gov
>> > >
>> > >
>> > >
>> > >
>> > > Katherine
>> > > McNeill-Harman
>> > > <mcneillh at mit.edu
>> >To
>> > > > ddi-adg at icpsr.umich.edu
>> > > Sent by:
>> >cc
>> > > ddi-adg-bounces at i
>> > > cpsr.umich.edu
>> >Subject
>> > > [DDI-ADG] standards for
>>frequency
>> > > (or lack thereof)
>> > > 08/02/2005 03:28
>> > > PM
>> > >
>> > >
>> > >
>> > >
>> > >
>> > >
>> > >
>> > >
>> > >Hi everyone,
>> > >
>> > >I've looked around fruitlessly for any standard list of frequencies.
I
>> > >looked at the ISO web site, but wasn't able to navigate to anything
>> >helpful
>> > >
>> > >(maybe someone else is more familiar). I did various web searches,
to
>>no
>> > >avail. I looked on the web sites and in reference books for bodies
that
>>I
>> > >know collect/publish data at certain frequencies (e.g. IMF, World
Bank).
>> > >I even looked in the dictionaries in which I found the other
>>definitions,
>> > >and several economic dictionaries.
>> > >
>> > >None of the sources above define/list these terms in any way. I'm
>> > >beginning to wonder if terms may be too taken-for-granted to be
listed
>>in
>> >a
>> > >
>> > >structured way.
>> > >
>> > >However, if nothing else, we could just go to a regular English
>>dictionary
>> > >for definitions of the terms. For example, my Websters Collegiate
>> > >Dictionary defines quarterly as "recurring, issued, or spaced at
3-month
>> > >intervals." This was the best that I found anywhere. This wouldn't
>>take
>> > >care of our need for a list from which to work in the first place,
but I
>> > >expect we'd be able to brainstorm most. (and then the "other" option
>>would
>> > >cover the rest)
>> > >
>> > >Maybe others have other ideas or techniques they could try, but I
think
>> >I'm
>> > >
>> > >stuck.
>> > >
>> > >Kate
>> > >
>> > >___________________________________________
>> > >Katherine McNeill-Harman
>> > >Data Services Librarian
>> > >Dewey Library for Management and Social Sciences
>> > >Massachusetts Institute of Technology
>> > >77 Massachusetts Avenue, E53-100
>> > >Cambridge, MA 02139
>> > >mcneillh at mit.edu
>> > >617-253-0787
>> > >
>> > >_______________________________________________
>> > >DDI-ADG mailing list
>> > >DDI-ADG at icpsr.umich.edu
>> > >http://www.icpsr.umich.edu/mailman/listinfo/ddi-adg
>> > >
>> > >_______________________________________________
>> > >DDI-ADG mailing list
>> > >DDI-ADG at icpsr.umich.edu
>> > >http://www.icpsr.umich.edu/mailman/listinfo/ddi-adg
>> >
>> >Mary Vardigan
>> >Director, Collection Delivery
>> >Inter-university Consortium for Political and Social Research (ICPSR)
>> >University of Michigan
>> >P.O. Box 1248, Ann Arbor, MI 48106-1248
>> >Phone: 734-615-7908
>> >Fax: 734-647-8200
>> >www.icpsr.umich.edu
>>
>>Mary Vardigan
>>Director, Collection Delivery
>>Inter-university Consortium for Political and Social Research (ICPSR)
>>University of Michigan
>>P.O. Box 1248, Ann Arbor, MI 48106-1248
>>Phone: 734-615-7908
>>Fax: 734-647-8200
>>www.icpsr.umich.edu
>>
>>_______________________________________________
>>DDI-ADG mailing list
>>DDI-ADG at icpsr.umich.edu
>>http://www.icpsr.umich.edu/mailman/listinfo/ddi-adg
>
>Mary Vardigan
>Director, Collection Delivery
>Inter-university Consortium for Political and Social Research (ICPSR)
>University of Michigan
>P.O. Box 1248, Ann Arbor, MI 48106-1248
>Phone: 734-615-7908
>Fax: 734-647-8200
>www.icpsr.umich.edu
>_______________________________________________
>DDI-ADG mailing list
>DDI-ADG at icpsr.umich.edu
>http://www.icpsr.umich.edu/mailman/listinfo/ddi-adg
___________________________________________
Katherine McNeill-Harman
Data Services Librarian
Dewey Library for Management and Social Sciences
Massachusetts Institute of Technology
77 Massachusetts Avenue, E53-100
Cambridge, MA 02139
mcneillh at mit.edu
617-253-0787
More information about the DDI-ADG
mailing list