goalie is an R package designed to provide a simple interface for extracting data from the United Nations Statistics Divisions (UNSD) Sustainable Development Goals (SDG) API. The package allows for exploration of SDG data contained within, including available dimensions/attributes and geographic coverage, while providing a simple tool to extract all data quickly into R.
sdg_overview(),sdg_targets(), andsdg_indicators()provide data frames covering the goals, targets, and indicators of the SDGs.sdg_dimensions()andsdg_attributes()provides a data frame of dimensions and attributes available for a specific SDG or series of data.sdg_series()provides a list of all data series available in the SDG database.sdg_geoareas()provides an overview of geographic areas covered by data in the SDG database, or that have data for a specific SDG goal, target, indicator, or series.sdg_geoarea_data()lists the SDG data available for a specific geographic area.sdg_data()extracts data for a selection of series in the GHO and outputs all results in a single data frame.
The interface is designed to be as simple as possible, only requiring input of the code of an indicator to extract it. However, at request, more detailed implementation of the SDG API’s systems can be implemented. Please provide any requests through the Github issues page.
goalie can be installed using remotes::install_github("gpw13/goalie")
To begin, we can use gho_overview() to begin to explore all data
available in the GHO.
library(goalie)
sdg_overview()
#> # A tibble: 14,199 x 12
#> goal goal_title goal_description target_description target_title target
#> <chr> <chr> <chr> <chr> <chr> <chr>
#> 1 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> 2 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> 3 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> 4 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> 5 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> # … with 14,194 more rows, and 6 more variables: indicator_description <chr>,
#> # indicator_tier <chr>, indicator <chr>, series_description <chr>,
#> # series_release <chr>, series <chr>If we want the data for SI_POV_DAY1, we could now just quickly access
the data frame using sdg_data().
sdg_data("SI_POV_DAY1")
#> # A tibble: 3,033 x 23
#> Goal Target Indicator SeriesCode SeriesDescription GeoAreaCode GeoAreaName
#> <dbl> <dbl> <chr> <chr> <chr> <dbl> <chr>
#> 1 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 2 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 3 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 4 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 5 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> # … with 3,028 more rows, and 16 more variables: TimePeriod <dbl>, Value <dbl>,
#> # Time_Detail <dbl>, TimeCoverage <lgl>, UpperBound <lgl>, LowerBound <lgl>,
#> # BasePeriod <lgl>, Source <chr>, GeoInfoUrl <lgl>, FootNote <chr>,
#> # Age <lgl>, Location <lgl>, Nature <chr>, Reporting_Type <chr>, Sex <lgl>,
#> # Units <chr>From here, standard methods of data manipulation (e.g. base R, the
tidyverse) could be used to select variables, filter rows, and explore
the data. However, we can also continue to explore other aspects of the
SDG database. For instance, if we wanted to see the dimensions and
attributes of SI_POV_DAY1, we can easily access that.
sdg_dimensions(series = "SI_POV_DAY1")
#> # A tibble: 142 x 4
#> id code description sdmx
#> <chr> <chr> <chr> <chr>
#> 1 Age <1M under 1 month old M0
#> 2 Age <1Y under 1 year old Y0
#> 3 Age <5Y under 5 years old Y0T4
#> 4 Age <15Y under 15 years old Y0T14
#> 5 Age <18Y under 18 years old Y0T17
#> # … with 137 more rowssdg_attributes(series = "SI_POV_DAY1")
#> # A tibble: 8 x 4
#> id code description sdmx
#> <chr> <chr> <chr> <chr>
#> 1 Nature C Country data C
#> 2 Nature CA Country adjusted data CA
#> 3 Nature E Estimated data E
#> 4 Nature G Global monitoring data G
#> 5 Nature M Modeled data M
#> # … with 3 more rowsLet’s say we want to get data for a specific country, then we could look up the M49 code using the table available through the API.
sdg_geoareas()
#> # A tibble: 390 x 2
#> geoAreaCode geoAreaName
#> <chr> <chr>
#> 1 4 Afghanistan
#> 2 248 Åland Islands
#> 3 8 Albania
#> 4 12 Algeria
#> 5 16 American Samoa
#> # … with 385 more rowsWe can then even check what data is available for a specific country, say Angola.
sdg_geoarea_data(24)
#> # A tibble: 454 x 12
#> goal goal_title goal_description target_description target_title target
#> <chr> <chr> <chr> <chr> <chr> <chr>
#> 1 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> 2 1 End poverty… Goal 1 calls for … By 2030, eradicat… By 2030, erad… 1.1
#> 3 1 End poverty… Goal 1 calls for … By 2030, reduce a… By 2030, redu… 1.2
#> 4 1 End poverty… Goal 1 calls for … By 2030, reduce a… By 2030, redu… 1.2
#> 5 1 End poverty… Goal 1 calls for … By 2030, reduce a… By 2030, redu… 1.2
#> # … with 449 more rows, and 6 more variables: indicator_description <chr>,
#> # indicator_tier <chr>, indicator <chr>, series_description <chr>,
#> # series_release <chr>, series <chr>And we can get data from the SDG for multiple series in one call, with the output data frames already merged together.
sdg_data(c("SI_POV_DAY1", "SI_POV_EMP1", "SI_POV_NAHC"))
#> # A tibble: 27,283 x 23
#> Goal Target Indicator SeriesCode SeriesDescription GeoAreaCode GeoAreaName
#> <dbl> <dbl> <chr> <chr> <chr> <dbl> <chr>
#> 1 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 2 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 3 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 4 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> 5 1 1.1 1.1.1 SI_POV_DAY1 Proportion of popu… 1 World
#> # … with 27,278 more rows, and 16 more variables: TimePeriod <dbl>,
#> # Value <dbl>, Time_Detail <dbl>, TimeCoverage <lgl>, UpperBound <lgl>,
#> # LowerBound <lgl>, BasePeriod <lgl>, Source <chr>, GeoInfoUrl <lgl>,
#> # FootNote <chr>, Age <chr>, Location <chr>, Nature <chr>,
#> # Reporting_Type <chr>, Sex <chr>, Units <chr>Of course, the reality is that it’s likely easier for us to work outside
the OData filtering framework and directly in R, so here’s a final more
complex example using dplyr and stringr alongside goalie to
automatically download all indicators for Angola with the word “poverty”
in the series description (case insensitive), for the years 1990 to
2005.
library(dplyr)
library(stringr)
sdg_geoarea_data(24) %>%
filter(str_detect(str_to_lower(series_description), "poverty")) %>%
pull(series) %>%
sdg_data(area_codes = 24, 1990, 2005)
#> # A tibble: 61 x 23
#> Goal Target Indicator SeriesCode SeriesDescription GeoAreaCode GeoAreaName
#> <dbl> <chr> <chr> <chr> <chr> <dbl> <chr>
#> 1 1 1.1 1.1.1 SI_POV_EMP1 Employed populatio… 24 Angola
#> 2 1 1.1 1.1.1 SI_POV_EMP1 Employed populatio… 24 Angola
#> 3 1 1.1 1.1.1 SI_POV_EMP1 Employed populatio… 24 Angola
#> 4 1 1.1 1.1.1 SI_POV_EMP1 Employed populatio… 24 Angola
#> 5 1 1.1 1.1.1 SI_POV_EMP1 Employed populatio… 24 Angola
#> # … with 56 more rows, and 16 more variables: TimePeriod <dbl>, Value <dbl>,
#> # Time_Detail <dbl>, TimeCoverage <lgl>, UpperBound <lgl>, LowerBound <lgl>,
#> # BasePeriod <lgl>, Source <chr>, GeoInfoUrl <lgl>, FootNote <chr>,
#> # Age <chr>, Location <lgl>, Nature <chr>, Reporting_Type <chr>, Sex <chr>,
#> # Units <chr>And once we have that data, we can then filter, explore, and analyze the data with our standard R workflow, or even export the downloaded data to Excel or other analytical tools for further use.