Easily seasonally adjust a tibble or data.frame of long data

seasthedata will automatically seasonally adjust every column in your tibble while respecting groups. Your data must contain exactly one column of dates.

seasthedata(original, frequency = NULL, use_original = FALSE, ...)

Arguments

original	Your data, in original terms, to be seasonally adjusted.
frequency	(Optional) The frequency of your data. Accepted options are "quarter", "month", "day". If omitted, seasthedata will guess the frequency.
use_original	(default: FALSE) If the series cannot be seasonally adjusted, should the returned data have NAs for the relevant series (default) or the original data.
...	Options to be passed to X13. See the documentation for the seas function (https://cran.r-project.org/web/packages/seasonal/seasonal.pdf)

Details

seasthedata is a wrapper around the seasonal library (https://github.com/christophsax/seasonal)

Examples

library(dplyr)
#> 
#> Attaching package: ‘dplyr’
#> The following objects are masked from ‘package:stats’:
#> 
#>     filter, lag
#> The following objects are masked from ‘package:base’:
#> 
#>     intersect, setdiff, setequal, union
library(tibble)
ungrouped_data <- tibble(dates = seq.Date(from = as.Date("1949-01-01"),
                         by = "month", length.out = 144),
                         y = as.vector(AirPassengers))
seasthedata(ungrouped_data)
#> Frequency not supplied, have determined it to be month
#> # A tibble: 144 × 2
#>    dates          y
#>    <date>     <dbl>
#>  1 1949-01-01  123.
#>  2 1949-02-01  125.
#>  3 1949-03-01  125.
#>  4 1949-04-01  128.
#>  5 1949-05-01  127.
#>  6 1949-06-01  126.
#>  7 1949-07-01  125.
#>  8 1949-08-01  127.
#>  9 1949-09-01  129.
#> 10 1949-10-01  129.
#> # … with 134 more rows

grouped_data <- bind_rows(mutate(ungrouped_data, group = "A"),
                          mutate(ungrouped_data, group = "B"))
grouped_data <- group_by(grouped_data, group)
seasthedata(grouped_data)
#> Frequency not supplied, have determined it to be month
#> # A tibble: 288 × 3
#>    dates          y group
#>    <date>     <dbl> <chr>
#>  1 1949-01-01  123. A    
#>  2 1949-02-01  125. A    
#>  3 1949-03-01  125. A    
#>  4 1949-04-01  128. A    
#>  5 1949-05-01  127. A    
#>  6 1949-06-01  126. A    
#>  7 1949-07-01  125. A    
#>  8 1949-08-01  127. A    
#>  9 1949-09-01  129. A    
#> 10 1949-10-01  129. A    
#> # … with 278 more rows