dmolitor / socr

Minimal client for SOCcer API.

socR

The goal of socR is to query the SOCcer API.

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("dmolitor/socR")

Example

This is a basic example of how to use the package:

library(socR)

Using the built-in data set we can query the first three jobs and see how long it takes. The first three job postings:

head(job_desc, 3)
#> # A tibble: 3 x 3
#>      job_id title                                         description           
#>       <dbl> <chr>                                         <chr>                 
#> 1 140906257 deli clerk customer service                   polite prompt knowled~
#> 2 140974158 roll off truck driver class b cdl 3 000 bonus operate manual automa~
#> 3 140923731 field engineer 3                              perform advanced trou~

Now, using SOCcer to return the top three SOC matches for each:

system.time({
  jobs <- soccer(
    job_title = job_desc$title[1:3],
    job_desc = job_desc$description[1:3],
    job_id = job_desc$job_id[1:3],
    n_results = 3,
    req_per_min = 12
  )
})
#>    user  system elapsed 
#>    0.05    0.06   11.00

Now we can view the top 3 results for each posting.

jobs
#> $problems
#> # A tibble: 0 x 0
#> 
#> $valid
#> # A tibble: 9 x 4
#>      job_id code    label                                                  score
#>       <dbl> <chr>   <chr>                                                  <dbl>
#> 1 140906257 43-4051 Customer Service Representatives                     0.202  
#> 2 140906257 35-2021 Food Preparation Workers                             0.107  
#> 3 140906257 35-3021 Combined Food Preparation and Serving Workers, Incl~ 0.0601 
#> 4 140974158 53-3032 Heavy and Tractor-Trailer Truck Drivers              0.969  
#> 5 140974158 53-3011 Ambulance Drivers and Attendants, Except Emergency ~ 0.00305
#> 6 140974158 55-3014 Artillery and Missile Crew Members                   0.00101
#> 7 140923731 43-9061 Office Clerks, General                               0.0196 
#> 8 140923731 17-2112 Industrial Engineers                                 0.00550
#> 9 140923731 17-2171 Petroleum Engineers                                  0.00550

The API responses are stored locally in a persistent cache to keep duplicate queries from hitting the API. As we can see, re-running the function returns cached results and finishes almost immediately.

system.time({
  jobs <- soccer(
    job_title = job_desc$title[1:3],
    job_desc = job_desc$description[1:3],
    job_id = job_desc$job_id[1:3],
    n_results = 3,
    req_per_min = 30
  )
})
#>    user  system elapsed 
#>    0.00    0.02    0.01

SOCcer Official

Daniel Russ, one of the creators of the SOCcer tool, has a similar package available here. This is the source of truth for the API query formatting, as there is no documentation currently.