The goal of sarbR is to provide an easy interface to the monthly release of the South African Reserve Bank Quarterly Bulletin data. Although the SARB has many other datasets, the Quarterly Bulletin makes up the bulk of the statistical release.
At the time of the publication of this package (April 2019), the SARB still has no easy way of accessing the data through an programmatic interface (API). The maintainer/s sarbR package aim to fill this role by curating the data automatically from the SARB, updating a dedicated database and giving access to the data through the simple sarb_code function.
Please be aware: the data is updated in an automated fashion as soon as a new release is published on the SARB's website.
-
This package is in no way endorsed or associated with the South African Reserve Bank (SARB).
-
Given that no formal API is provided, the maintainers of this package rely on their own infrastructure to collect, curate, clean, maintain an up to date database and give access through a securely built API.
-
Although the greatest care is taken to provide good data, the maintainer/s of this package take no responsibility in the service OR data being supplied.
The package is built around the main sarb_code function, with two helper functions. The first providing a way to search for an indicator and its meta information: search_indicator. The second function is to request an access token for the API: token_request.
To install the package
# Install development version from GitHub
remotes::install_github("Stellenbosch-Econometrics/sarbR")
library(sarbR)Because the database is maintained on a small server and the maintainer wishes to keep the server protected from malicious conduct, it was decided to implement requests on a token basis. This ensures some measure of security, a bit of analytics on request and hopefuly allows the small server to handle everyones request in an orderly manner.
To request a token from the service follow 2 simple steps:
- Step 1: request a token
token_request()
#> Successful token creation
#> $status
#> [1] 200
#>
#> $msg
#> [1] "Successful token creation"
#>
#> $info
#> address token
#> 122.232.34.123 f84sdsfsdce65eege7dbacd93ac3dc073e364- Step 2: add token to environment or options,
usethis::edit_r_environ(), or set it in your current session using:
options("sarb.token" = "f84sdsfsdce65eege7dbacd93ac3dc073e364")
Sys.setenv("sarb.token" = "f84sdsfsdce65eege7dbacd93ac3dc073e364")release_info()
#> [1] "Latest release info: 2020-12-15"all_codes()
#># A tibble: 4,316 x 1
#> code
#> <chr>
#> 1 KBP1000J
#> 2 KBP1000M
#> 3 KBP1005J
#> 4 KBP1005M
#> 5 KBP1006J
#> 6 KBP1006M
#> 7 KBP1007J
#> 8 KBP1007M
#> 9 KBP1008J
#>10 KBP1008M
#># … with 4,306 more rowsThis way you will not need to send the token along when making a request for data
The sarb_code function is the main function of the package. This functions interacts with a plumber API that has access to a MySQL database with the latest quarterly data.
To use request the data, the whole code needs to be provided. As an example, if you would like to reques Expenditure on gross domestic product at Current prices at a yearly frequency, you need to request KBP6045J
sarb_code(code = "KBP6045J")Or if the the token is not set as in the above mentioned manner (in the .Renviron), then it can be passed along to the function
sarb_code(code = "KBP6045J", token = "f84sdsfsdce65eege7dbacd93ac3dc073e364")The SARB provides various series with specific codes linked to a series: KBPXXXX. It can be difficult to know which series you are looking.
To help you find a specific frequency of a series (Monthly, Quarterly, Yearly) or just look in general for a series on say, GDP, we can use the search_indicator function. This function is a convenience wrapper around the indicator_codes dataset
Using the description to look for specific series:
search_indicator(description = "Expenditure")
#> # A tibble: 6 x 7
#> time_series description_and_vers… series version_description frequency unit_of_measure code
#> <chr> <S3: glue> <chr> <chr> <chr> <chr> <S3: >
#> 1 KBP6045 Expenditure on gross… C Constant 2010 prices K1 RMILL KBP60…
#> 2 KBP6045 Expenditure on gross… D Constant 2010 prices. S… K1 RMILL KBP60…
#> 3 KBP6045 Expenditure on gross… J Current prices J1 RMILL KBP60…
#> 4 KBP6045 Expenditure on gross… K Current prices K1 RMILL KBP60…
#> 5 KBP6045 Expenditure on gross… L Current prices. Seasona… K1 RMILL KBP60…
#> 6 KBP6045 Expenditure on gross… Y Constant 2010 prices J1 RMILL KBP60…Look up the time series:
search_indicator(time_series = "KBP1000")
#> # A tibble: 2 x 7
#> time_series description_and_version series version_descript… frequency unit_of_measure code
#> <chr> <S3: glue> <chr> <chr> <chr> <chr> <S3: >
#> 1 KBP1000 South African Reserve Bank li… J NA J1 RMILL KBP10…
#> 2 KBP1000 South African Reserve Bank li… M NA M1 RMILL KBP10…You can also look up a specific series if you know the exact code:
search_indicator(code = "KBP1000J")
#> # A tibble: 2 x 7
#> time_series description_and_version series version_descript… frequency unit_of_measure code
#> <chr> <S3: glue> <chr> <chr> <chr> <chr> <S3: >
#> 1 KBP1000 South African Reserve Bank li… J NA J1 RMILL KBP10…The package also includes data that helps you understand the frequency column used in the indicator_codes dataset
sarbR::frequency_description
#> Description Frequency
#> 1 D6 Daily - Monday to Saturday
#> 2 W3 Weekly as on Wednesday
#> 3 W6 Weekly - Monday to Saturday
#> 4 M1 Monthly
#> 5 K1 Quarterly
#> 6 J1 Annually - 1 January to 31 December
#> 7 J2 Annually - 1 April to 31 MarchIf you want to make a specific dataset available through the package, a request can be made through the issues page:
https://github.com/Stellenbosch-Econometrics/sarbR/issues