What are a mnemonic, concept code and geography code? How do they identify Data Buffet time series, singly or in related groups?
Every time series in Data Buffet is identified with a unique alphanumeric name called a mnemonic (pronounced "nuh-mon-ick"). This consists of two parts, separated by a period: a concept code (or "concept") denoting the indicator measured, and a geography code (or "geo") specifying the place of measurement. For instance, "ET.US" (pronounced "E T dot U S") is nonfarm employment for the United States from the BLS CES report: Also:
- XMBAD30.PA = Home mortgages delinquent by 30-59 days - for Pennsylvania
- RHHOLDA.PA001 = Estimated population - for Pennsylvania - Adams County
- NALEGDPUQ.INOR = Gross domestic product - for Norway
Related indicators have concept codes with similar structures, so that a "wild card expression" in a basket can retrieve them in bulk. For instance, if the dataset is reported by NAICS industry, we embed the official numeric code. For example:
- RYPEW23Q.IUSA_DMCHI = Wages and salaries for construction (NAICS 23) - for Chicago MSA
- RYPEW236Q.IUSA_DMCHI = For construction of buildings (NAICS 236)
- RYPEW2362Q.IUSA_DMCHI = For construction of non-residential buildings (NAICS 2362)
- RYPEW23?Q.IUSA_DMCHI = All industries within construction
We define geo codes for countries, subnational areas, and supranational aggregates. Our geo codes are regularized but not necessarily recognizable by name, in part because an area may have very different names under different languages. In some cases, we are able to embed official geographic code numbers within our geo code.
Many (but not all) geo types are distinguishable by wild card. For example:
- LWEMEDPA.IGBR = Median annual income, for the U.K.
- LWEMEDPA.IGBR_SCTL = For Scotland
- LWEMEDPA.IGBR_WALS = For Wales
- LWEMEDPA.IGBR_^^^^ = All U.K. NUTS 1 areas
Keep in mind:
We design mnemonics to have some regularity but they cannot, in general, be guessed in a vacuum; hence, for purposes of first retrieving series, we recommend that you browse the Data Buffet catalog. Once you have collected a basket, mnemonics are unambiguous when you seek support. Given one mnemonic, we can suggest variations and wild card expressions to retrieve related series.
A single series may have multiple mnemonics, including a preferred canonical mnemonic and additional aliases for backwards compatibility, brevity, or cross-reference. Aliases are listed in Mnemonic 411.
Ideally, every series sharing a concept code would describe the same measurement, and every comparable measurement would be named with the same concept. Given the scope of our repository, this is only approximately true.
Sometimes the "same" variable is several distinct series with distinct mnemonics. For instance, there are many alternative house prices indexes published by different agencies. Or, the same measurement is reprised in multiple reports, but on different schedules.
The "same" area under a given commonly-used name may be described with multiple geo codes, because areas are not static; an individual area may be redefined, or may be part of larger classification in which many (but not all) areas are redefined.