Skip to content

Latest commit

 

History

History
263 lines (195 loc) · 11.3 KB

README.md

File metadata and controls

263 lines (195 loc) · 11.3 KB

blscrapeR

R build status CRAN_Status_Badge Project Status: Active - The project has reached a stable, usable state and is being actively developed.

Designed to be a tidy API wrapper for the Bureau of Labor Statistics (BLS.) The package has additional functions to help parse, analyze and visualize the data. The package utilizes "tidyverse" concepts for internal functionality and encourages the use of those concepts with the output data.

Install

  • Stable version from CRAN:
install.packages("blscrapeR")
  • The latest development version from GitHub:
devtools::install_github("keberwein/blscrapeR")

Before getting started, you’ll probably want to head over to the BLS and get set up with an API key. While an API key is not required to use the package, the query limits are much higher if you have a key and you’ll have access to more data. Plus, it’s free (as in beer), so why not?

Basic Usage

For “quick and dirty” type of analysis, the package has some quick functions that will pull metrics from the API without series numbers. These quick functions include unemployment, employment, and civilian labor force on a national level.

library(blscrapeR)
# Grab the Unemployment Rate (U-3) 
df <- quick_unemp_rate()
head(df, 5)
#> # A tibble: 5 x 6
#>    year    period periodName value footnotes  seriesID
#>   <dbl>    <list>     <list> <dbl>    <list>    <list>
#> 1  2017 <chr [1]>  <chr [1]>   4.1 <chr [1]> <chr [1]>
#> 2  2017 <chr [1]>  <chr [1]>   4.2 <chr [1]> <chr [1]>
#> 3  2017 <chr [1]>  <chr [1]>   4.4 <chr [1]> <chr [1]>
#> 4  2017 <chr [1]>  <chr [1]>   4.3 <chr [1]> <chr [1]>
#> 5  2017 <chr [1]>  <chr [1]>   4.4 <chr [1]> <chr [1]>

Search BLS IDs

Some knowledge of BLS ids are needed to query the API. The package includes a "fuzzy search" function to help find these ids. There are currently more than 75,000 series ids in the package's internal data set, series_ids. While these aren't all the series ids the BLS offers, it contains the most popular. The BLS Data Finder is another good resource for finding series ids, that may not be in the internal data set.

library(blscrapeR)
# Find series ids relating to the total labor force in LA.
ids <- search_ids(keyword = c("Labor Force", "Los Angeles"))
head(ids)
#> # A tibble: 6 x 4
#>                                                                  series_title
#>                                                                         <chr>
#> 1 Labor Force: Balance Of California, State Less Los Angeles-Long Beach-Glend
#> 2  Labor Force: Los Angeles-Long Beach-Glendale, Ca Metropolitan Division (S)
#> 3 Labor Force: Balance Of California, State Less Los Angeles-Long Beach-Glend
#> 4       Labor Force: Los Angeles-Long Beach, Ca Combined Statistical Area (U)
#> 5                                     Labor Force: Los Angeles County, Ca (U)
#> 6                                       Labor Force: Los Angeles City, Ca (U)
#> # ... with 3 more variables: series_id <chr>, seasonal <chr>,
#> #   periodicity_code <chr>
library(blscrapeR)
# Find series ids relating to median weekly earnings of women software developers.
ids <- search_ids(keyword = c("Earnings", "Software", "Women"))
head(ids)
#> # A tibble: 1 x 4
#>                                                                  series_title
#>                                                                         <chr>
#> 1 (Unadj)- Median Usual Weekly Earnings (Second Quartile), Employed Full Time
#> # ... with 3 more variables: series_id <chr>, seasonal <chr>,
#> #   periodicity_code <chr>

API Keys

You should consider getting an API key form the BLS. The package has a function to install your key in your .Renviron so you’ll only have to worry about it once. Plus, it will add extra security by not having your key hard-coded in your scripts for all the world to see.

From the BLS:

Service Version 2.0 (Registered) Version 1.0 (Unregistered)
Daily query limit 500 25
Series per query limit 50 25
Years per query limit 20 10
Net/Percent Changes Yes No
Optional annual averages Yes No
Series descriptions Yes No

Download Multiple BLS Series at Once

library(blscrapeR)

# Grab several data sets from the BLS at onece.
# NOTE on series IDs: 
# EMPLOYMENT LEVEL - Civilian labor force - LNS12000000
# UNEMPLOYMENT LEVEL - Civilian labor force - LNS13000000
# UNEMPLOYMENT RATE - Civilian labor force - LNS14000000
df <- bls_api(c("LNS12000000", "LNS13000000", "LNS14000000"),
              startyear = 2008, endyear = 2017, Sys.getenv("BLS_KEY")) %>%
    # Add time-series dates
    dateCast()
# Plot employment level
library(ggplot2)
gg1200 <- subset(df, seriesID=="LNS12000000")
library(ggplot2)
ggplot(gg1200, aes(x=date, y=value)) +
    geom_line() +
    labs(title = "Employment Level - Civ. Labor Force")

Median Weekly Earnings

library(blscrapeR)
library(tidyverse)
# Median Usual Weekly Earnings by Occupation, Unadjusted Second Quartile.
# In current dollars
df <- bls_api(c("LEU0254530800", "LEU0254530600"), startyear = 2000, endyear = 2016, registrationKey = Sys.getenv("BLS_KEY")) %>%
    spread(seriesID, value) %>% dateCast()
# A little help from ggplot2!
library(ggplot2)
ggplot(data = df, aes(x = date)) + 
    geom_line(aes(y = LEU0254530800, color = "Database Admins.")) +
    geom_line(aes(y = LEU0254530600, color = "Software Devs.")) + 
    labs(title = "Median Weekly Earnings by Occupation") + ylab("value") +
    theme(legend.position="top", plot.title = element_text(hjust = 0.5)) 

For more advanced usage, please see the package vignettes.

Inflation and Consumer Price Index (CPI)

Although there are many measures of inflation, the CPI's "Consumer Price Index for All Urban Consumers: All Items" is normally the headline inflation rate one would hear about on the news (see FRED).

Getting these data from the blscrapeR package is easy enough:

library(blscrapeR)
df <- bls_api("CUSR0000SA0")
head(df)

Due to the limitations of the API, we are only able to gather twenty years of data per request. However the formula for calculating inflation is based on the 1980 dollar, so the data from the API aren't sufficient.

The package includes a function that collects information form the CPI beginning at 1947 and calculates inflation.

To find out the value of a January 2015 dollar in January 2023, we just make a simple function call. Looking at the adj_dollar_value column. We can see that the value of a 2015 dollar in 2023 was approximately $1.32.

df <- inflation_adjust("2015-01-01") %>%
    arrange(desc(date))
head(df)

library(blscrapeR)
# A tibble: 6 × 7
  date       period year  value base_date  adj_dollar_value month_ovr_month_pct_change
  <date>     <chr>  <chr> <dbl> <chr>                 <dbl>                      <dbl>
1 2024-02-01 M02    2024   310. 2015-01-01             1.33                      0.619
2 2024-01-01 M01    2024   308. 2015-01-01             1.32                     -0.105
3 2023-12-01 M12    2023   307. 2015-01-01             1.31                     -0.415
4 2023-12-01 M12    2023   309. 2015-01-01             1.32                      0.651
5 2023-11-01 M11    2023   307. 2015-01-01             1.31                     -0.156
6 2023-11-01 M11    2023   308. 2015-01-01             1.32                      0.317


If we want to check our results, we can head over to the CPI Inflation Calculator on the BLS website.

Annual Inflation Percentage Increase

library(blscrapeR)
library(ggplot2)

ggplot(data = df, aes(x = date)) + 
    geom_line(aes(y = adj_dollar_value, color = "2015 Adjusted Dollar Value")) + 
    labs(title = "Inflation Since 2015") + ylab("2015 Adjusted Dollar Value") +
    theme(legend.position="top", plot.title = element_text(hjust = 0.5)) 


ggplot(data = df, aes(x = date)) + 
    geom_line(aes(y = month_ovr_month_pct_change, color = "MoM Pct Change")) + 
    labs(title = "Month over Month Inflation Pct Change") + ylab("MoM Pct Change") +
    theme(legend.position="top", plot.title = element_text(hjust = 0.5)) 



CPI: Tracking Escalation

Another typical use of the CPI is to determine price escalation. This is especially common in escalation contracts. While there are many different ways one could calculate escalation below is a simple example. Note: the BLS recommends using non-seasonally adjusted data for escalation calculations.

Suppose we want the price escalation of $100 investment we made in January 2014 to February 2015:

Disclaimer: Escalation is normally formulated by lawyers and bankers, the author(s) of this package are neither, so the above should only be considered a code example.

library(blscrapeR)
library(dplyr)
df <- bls_api("CUSR0000SA0",
              startyear = 2014, endyear = 2015)
head(df)

# A tibble: 6 × 6
   year period periodName value footnotes seriesID   
  <dbl> <chr>  <chr>      <dbl> <chr>     <chr>      
1  2015 M12    December    238. ""        CUSR0000SA0
2  2015 M11    November    238. ""        CUSR0000SA0
3  2015 M10    October     238. ""        CUSR0000SA0
4  2015 M09    September   237. ""        CUSR0000SA0
5  2015 M08    August      238. ""        CUSR0000SA0
6  2015 M07    July        238. ""        CUSR0000SA0

# Set base value.
base_value <- 100

# Get CPI from base period (January 2014).
base_cpi <- subset(df, year==2014 & periodName=="January", select = "value")

# Get the CPI for the new period (February 2015).
new_cpi <- subset(df, year==2015 & periodName=="February", select = "value")

# Calculate the updated value of our $100 investment.
round((base_value / base_cpi) * new_cpi, 2)
   value
1 100.02

# Huzzah! We made 2 cents on our $100 investment.