Data (New): Difference between revisions

From Pardee Wiki
Jump to navigation Jump to search
(Created page with "== '''Database Files in IFs''' == === Historical Databases === International Futures (IFs) modeling application contains two large historical databases, IFsHistSeries & IFsHistSeriesDyadic. The former provides monadic information (values per entity), and the latter provides bilateral information (values per entity-pair). These databases consist of a large collection of time-series tables. Hence, each data table is called a data series in the IFs system. These data serie...")
 
mNo edit summary
Line 2: Line 2:


=== Historical Databases ===
=== Historical Databases ===
International Futures (IFs) modeling application contains two large historical databases, IFsHistSeries & IFsHistSeriesDyadic. The former provides monadic information (values per entity), and the latter provides bilateral information (values per entity-pair). These databases consist of a large collection of time-series tables. Hence, each data table is called a data series in the IFs system. These data series cover a large number of variables on IFs model issue areas like population, education, health, agriculture, energy, economics, environment, infrastructure, governance and international politics.
The International Futures (IFs) modeling system relies on two major large historical databases: '''IFsHistSeries''' and '''IFsHistSeriesDyadic'''.


At present IFs data tables contain data for 188 entities (countries/regions). Temporal coverage, in most cases, start in 1960.
* '''IFsHistSeries''' contains data for individual countries and regions (''monadic data'').
* '''IFsHistSeriesDyadic''' holds data on relationships between country pairs (''bilateral data'').


The model uses these data tables to,
Both databases store a vast collection of time-series data tables covering areas like population, education, health, agriculture, energy, economics, the environment, infrastructure, governance, and international politics. Hence, each data table is also referred to as a data series. Currently, these data series include '''188 entities (countries/regions)''', with most data series going back to '''1960'''.


1.     Initialize values for IFs variables in the base year. These tables are also called preprocessors series.
The IFs model uses these data series in several ways:


2.     Build models to analyze and use the relationships between variables. These tables are also called series for analytical functions.
# '''Initializing model variables'''—these series act as ''preprocessors'' that set the base-year values for the model.
# '''Building relationships between variables'''—helping IFs analyze connections between different variables, e.g., through a statistical model. These tables serve as ''Analytical Function'' Series.
# '''Displaying historical trends'''—providing users with visual insights into past patterns. These are also referred to as series for ''Historical Analog''.
# '''Incorporating external projections'''—allowing comparisons with forecasts from other organizations. Furthermore, some of these series can be made exogenous in IFs to change model projections.
# '''Helping users explore patterns and relationships'''—supporting research, analysis, and parameter estimation.


3.     Display historical trends in the IFs interface. These tables are also referred to as historical analog series.
For a deeper dive into the database structure, check out [a wiki link to IFsHistSeries & IFsHistSeriesDyadic].
 
4.     Provide projections from other agencies for comparisons and analysis.
 
5.     Enables model users to explore patterns, understand relationships and estimate parameters.
 
More details on the structure of historical databases can be found here.


=== Meta Data ===
=== Meta Data ===

Revision as of 21:34, 11 February 2025

Database Files in IFs

Historical Databases

The International Futures (IFs) modeling system relies on two major large historical databases: IFsHistSeries and IFsHistSeriesDyadic.

  • IFsHistSeries contains data for individual countries and regions (monadic data).
  • IFsHistSeriesDyadic holds data on relationships between country pairs (bilateral data).

Both databases store a vast collection of time-series data tables covering areas like population, education, health, agriculture, energy, economics, the environment, infrastructure, governance, and international politics. Hence, each data table is also referred to as a data series. Currently, these data series include 188 entities (countries/regions), with most data series going back to 1960.

The IFs model uses these data series in several ways:

  1. Initializing model variables—these series act as preprocessors that set the base-year values for the model.
  2. Building relationships between variables—helping IFs analyze connections between different variables, e.g., through a statistical model. These tables serve as Analytical Function Series.
  3. Displaying historical trends—providing users with visual insights into past patterns. These are also referred to as series for Historical Analog.
  4. Incorporating external projections—allowing comparisons with forecasts from other organizations. Furthermore, some of these series can be made exogenous in IFs to change model projections.
  5. Helping users explore patterns and relationships—supporting research, analysis, and parameter estimation.

For a deeper dive into the database structure, check out [a wiki link to IFsHistSeries & IFsHistSeriesDyadic].

Meta Data

All meta data (e.g., definitions, units, data source) are stored separately under the data dictionary files (DataDict.db or DataDictDyadic.db). For each data series in the historical databases, we have a reference (one row) in the data dictionary. IFs users can access this information directly through the database files, or through the IFs interface by clicking the Data Information tab when available.

Detailed description on the structure of data dictionary files can be found here.

Sing-Year Data and Parameters

In addition to historical databases. IFs sometimes relies on single-year data in its model initialization and projection. Single-year data are normally stored under the file IFs.db. Some model parameters and settings, such as country concordance tables, can also be found under the same file. You can find more information about tables in the IFs.db here.

Data Collection & Update

IFs databases are compiled, in most cases, from publicly available data sources, e.g., World Development Indicators, UNESCO Institute for Education database on education, FAO database on agricultural production, Correlates of War data on conflicts etc. When needed, new series will be added to the databases, while existing ones are updated by data sources on a regular basis.

Introductions on data collection can be traced using the list of data sources here.