In dataobservatory-eu/indicator: Tidy Statistical Indicators

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(indicators)

Indicator Table

indicator_code: Unique code for the indicator; the values are unique in indicator_code, geo, time, value, and unit.
unit: measurement unit or currency code
geo: geographical code of observation
time: Time of observation, as Date
value: Value of observation or indicator
year: Year of observation
month: Month of observation (often notional for annual indicators)
day: Day of observation
frequency: Frequency of observation, currently A as annual, Q as quarterly, M as monthly, D as daily, or unknown.
estimate: See below.
db_source: Database source code.

Estimates

The estimate label describes the statistical technique used to impute missing values, or fix geographical coding.

actual: The actual observation value at source, for example, by Eurostat.
missing: Missing data, which is not filled by locf, nocb, interpolate, forecast, backcast or impute.
locf: last observation carry forward
nocb: next observation carry backward
interpolate: interpolation in time between known values
forecast: forecast with a given model.
backcast: backcast with a given model.
impute: An imputation technique was used that is not based on time series, but on a crossselection, or geographical, or other information. This is likely to be rather specific to the actual indicator.

The recode means that the regional codes needed to be fixed.

Methods

The method label describes the statistical technique used to give the parameters oft the estimate, if the estimate is parametric.

The recode means that the regional codes needed to be fixed.

Metadata

indicator_code: the same as in value table.
title_at_source: The name, title, or other human understandable description of the original source.
description_indicator: A short textual description of the indicator. [A longer should be generated in html somewhere.]
db_source_code: A unique source code that contains both indicator_code and source for easier programatic updating.
frequency: Frequency of the indicator observations, currently Annual, Quarterly, Monthly, Daily, or unknown.
data_start: The minimum value of actual observations. Further values may be backcasted.
data_end: The last value of actual observations. Further values may be forecasted.
last_update_data_source: Last update at the data source.
last_structure_change: Last structural change at source.
last_update_data: Last time the observatory updated the data, including various imputations, forecasts and backcasts.
actual: The number of actual observation values at source, for example, by Eurostat.
missing: Number of missing data, which is not filled by locf, nocb, interpolate, forecast, backcaskt or impute.
locf: number of last observation carry forward imputated values
nocb: number of next observation carry backward imputated values
interpolate: number of interpolated values
forecast: number of values forecasted with a given model.
backcast: number of values backcasted with a given model. -impute: number of values filled with any other imputation method that is not based on time series, but cross-selectional, geographical or other information.