The goal of this package is to provide functions to handle datasets with HXL tags. It is currently work in progress. If you have any ideas on how to further develop the package, please feel free to open an issue or a pull request.

HXL is a different kind of data standard, designed to improve information sharing during a humanitarian crisis without adding extra reporting burdens. hxlstandard.org

Install

To install the current development version use devtools:

devtools::install_github("dirkschumacher/rhxl")

API by Example

Read in the aiports of Viet Nam.

data_url <- "http://ourairports.com/countries/VN/airports.hxl"
hxl_data <- as_hxl(read.csv(data_url))
head(hxl_data)
#> # A tibble: 6 x 20
#>      id ident type     name        latitude_deg longitude_deg elevation_ft
#>   <int> <chr> <chr>    <chr>              <dbl>         <dbl>        <int>
#> 1 26708 VVTS  large_a… Tan Son Nh…         10.8           107           33
#> 2 26700 VVNB  large_a… Noi Bai In…         21.2           106           39
#> 3 26697 VVDN  large_a… Da Nang In…         16.0           108           33
#> 4 26705 VVPQ  medium_… Phu Quoc I…         10.2           104           37
#> 5 26693 VVCR  medium_… Cam Ranh A…         12.0           109           40
#> 6 26702 VVPB  medium_… Phu Bai Ai…         16.4           108           48
#> # ... with 13 more variables: continent <chr>, iso_country <chr>,
#> #   iso_region <chr>, municipality <chr>, scheduled_service <int>,
#> #   gps_code <chr>, iata_code <chr>, local_code <chr>, home_link <chr>,
#> #   wikipedia_link <chr>, keywords <chr>, score <int>, last_updated <dttm>

You can get the schema as a tidy long table:

hxl_schema(hxl_data)
#> # A tibble: 35 x 3
#>    tag   attribute column_idx
#>    <chr> <chr>          <int>
#>  1 meta  id                 1
#>  2 meta  code               2
#>  3 loc   airport            3
#>  4 loc   type               3
#>  5 loc   airport            4
#>  6 loc   name               4
#>  7 geo   lat                5
#>  8 geo   lon                6
#>  9 geo   elevation          7
#> 10 geo   ft                 7
#> # ... with 25 more rows

Or as a character vector in the order of the columns:

hxl_schema_chr(hxl_data)
#>  [1] "#meta +id"                  "#meta +code"               
#>  [3] "#loc +airport +type"        "#loc +airport +name"       
#>  [5] "#geo +lat"                  "#geo +lon"                 
#>  [7] "#geo +elevation +ft"        "#region +continent +code"  
#>  [9] "#country +code +iso2"       "#adm1 +code +iso"          
#> [11] "#loc +municipality +name"   "#status +scheduled"        
#> [13] "#loc +airport +code +gps"   "#loc +airport +code +iata" 
#> [15] "#loc +airport +code +local" "#meta +url +airport"       
#> [17] "#meta +url +wikipedia"      "#meta +keywords"           
#> [19] "#meta +score"               "#date +updated"

We can also test, if a dataset supports a schema. For example, we could test if the dataset has lat/lng coordinates.

hxl_validate(hxl_data, c("#geo +lat", "#geo +lon"))
#> [1] TRUE

With this information you could for example write a function that can automatically display a dataset on a map - without much configuration by the user. It also makes sharing of information easier.

In addition you can select certain columns based on tags:

hxl_select(hxl_data, c("#geo +lat", "#geo +lon"))
#> # A tibble: 47 x 2
#>    latitude_deg longitude_deg
#>  *        <dbl>         <dbl>
#>  1        10.8            107
#>  2        21.2            106
#>  3        16.0            108
#>  4        10.2            104
#>  5        12.0            109
#>  6        16.4            108
#>  7        12.2            109
#>  8         8.73           107
#>  9        10.1            106
#> 10        20.8            107
#> # ... with 37 more rows