Create codes to identify each unique combination of hierarchical levels in a reference dataset

Create codes to identify each unique combination of hierarchical levels in a reference dataset. Codes may be integer-based (function hcodes_int) or string-based (hcodes_str). Integer-based codes reflect the alphabetical ranking of each level within the next-highest level. They are constant-width and may optionally be prefixed with any given string. String-based codes are created by pasting together the values of each hierarchical level with a given separator (with options for string standardization prior to collapsing).

Usage

hcodes_str(ref, pattern, by, sep = "__", std_fn = string_std)

hcodes_int(ref, pattern, by, prefix = "")

Arguments

ref: data.frame containing hierarchical columns with reference data
pattern: regex pattern to match the names of the hierarchical columns in ref (supply either pattern or by)
by: vector giving the names of the hierarchical columns in ref (supply either pattern or by)
sep: (only for hcodes_str) desired separator between levels in string-based codes (defaults to "__")
std_fn: (only for hcodes_str) Function to standardize input strings prior to creating codes. Defaults to string_std. Set to NULL to omit standardization. See also string_standardization.
prefix: (only for hcodes_int) character prefix for integer-based codes (defaults to "")

Value

A vector of codes

Examples

data(ne_ref)

# string-based codes
hcodes_str(ne_ref, pattern = "^adm")
#>  [1] "can"                             "usa"                            
#>  [3] "can__ontario"                    "usa__new_jersey"                
#>  [5] "usa__new_york"                   "usa__pennsylvania"              
#>  [7] "can__ontario__durham"            "can__ontario__halton"           
#>  [9] "can__ontario__peel"              "can__ontario__toronto"          
#> [11] "can__ontario__york"              "usa__new_jersey__bergen"        
#> [13] "usa__new_jersey__essex"          "usa__new_jersey__hudson"        
#> [15] "usa__new_jersey__middlesex"      "usa__new_jersey__monmouth"      
#> [17] "usa__new_york__jefferson"        "usa__new_york__bronx"           
#> [19] "usa__new_york__kings"            "usa__new_york__nassau"          
#> [21] "usa__new_york__new_york"         "usa__new_york__queens"          
#> [23] "usa__new_york__suffolk"          "usa__pennsylvania__allegheny"   
#> [25] "usa__pennsylvania__bucks"        "usa__pennsylvania__chester"     
#> [27] "usa__pennsylvania__delaware"     "usa__pennsylvania__jefferson"   
#> [29] "usa__pennsylvania__lancaster"    "usa__pennsylvania__philadelphia"
#> [31] "usa__pennsylvania__york"        

# integer-based codes
hcodes_int(ne_ref, pattern = "^adm")
#>  [1] "100" "200" "110" "210" "220" "230" "111" "112" "113" "114" "115" "211"
#> [13] "212" "213" "214" "215" "222" "221" "223" "224" "225" "226" "227" "231"
#> [25] "232" "233" "234" "235" "236" "237" "238"