R
Run advanced statistical analyses on Pacific data using the rsdmx package
Last updated
Run advanced statistical analyses on Pacific data using the rsdmx package
Last updated
PDH.stat is part of the rsdmx package, developed by Emmanuel Blondel, and contributors Matthieu Stigler and Eric Persson. Learn more about the original package here. It has been configured to include Pacific Data Hub's .Stat API as a default service provider.
These steps have been tested with R 4.0.2 on Windows 10.
Remove rsdmx if already installed: remove.packages("rsdmx")
Install devtools: install.packages("devtools")
Install rsdmx from the latest development version on Github: devtools::install_github("opensdmx/rsdmx")
This is a quick-start guide. Go here for the official documentation.
Load package: library(rsdmx)
See all service providers
Aside from PDH.stat, the original package offers connectivity with OECD, Eurostat and others. See all available service providers with the getSDMXServiceProviders() function.
See available dataflows from PDH.stat
To see the available PDH.stat dataflows (data sets), use the readSDMX() function, setting the providerId parameter to "PDH" and the resource parameter to "dataflow":
To return the available data set IDs and their English names, filter the dataframe:
Get all data for a dataflow
To retrieve a dataflow, provide the dataflow ID to the readSDMX() function in the flowref parameter, also setting the resource as "data".
For example, to connect to "Inflation Rates" dataflow, the ID is "DF_CPI" (as shown when retrieving all the dataflows for PDH.stat):
Get more specific data for a dataflow
Extra parameters can be supplied the readSDMX() function to retrieve a filtered view of the dataflow:
start is the desired start year (supplied as an integer)
end is the desired end year (supplied as an integer)
key controls a variety of filters, and by default it is set to "all" (retrieves all data). A further explanation is provided below.
The key parameter controls a different number of variables depending on the dataflow, including time period, country, currency and others. Each variable is selected with a code, and separated by a dot . Two dots .. indicates a "wildcard" (selects all available values). A plus + can allow multipled variables to be selected. Generally the time period comes first, A for "annual" or M for "month" (if the data is available at that level). Some examples:
For DF_CPI "Inflation Rates" dataflow, to get annual data from 2010-2015 for Cook Islands and Fiji:
The key is "A.CK+FJ.."
start is 2010 and end is 2015
The R code:
Given that the key variables can change depending on the dataflow, it can be easier to retrieve all data and then filter manually in R. Alternatively, use the Data Explorer to filter a dataset and then view the relevant API call and key as explained here.
