This seems like it would be a simple thing to do, but I am stumped.
I was using tidyverse material as a guide: here
I have a list of recession time periods, and I want to create a data frame as an output that lists every date and whether or not that date is in a recession. I would like to keep the solution in dplyr format.
Here is a reproducible example
library(lubridate)
library(tidyverse)
# Sample data set
my_df <-
structure(list(recession_start = structure(c(1400, 3652, 4199,
7486, 11382, 13848), class = "Date"), recession_end = structure(c(1885,
3834, 4687, 7729, 11627, 14396), class = "Date"), recession_interval = new("Interval",
.Data = c(41904000, 15724800, 42163200, 20995200, 21168000,
47347200), start = structure(c(120960000, 315532800, 362793600,
646790400, 983404800, 1196467200), tzone = "UTC", class = c("POSIXct",
"POSIXt")), tzone = "UTC")), row.names = c(NA, -6L), class = c("tbl_df",
"tbl", "data.frame"))
> my_df
# A tibble: 6 x 3
recession_start recession_end recession_interval
<date> <date> <Interval>
1 1973-11-01 1975-03-01 1973-11-01 UTC--1975-03-01 UTC
2 1980-01-01 1980-07-01 1980-01-01 UTC--1980-07-01 UTC
3 1981-07-01 1982-11-01 1981-07-01 UTC--1982-11-01 UTC
4 1990-07-01 1991-03-01 1990-07-01 UTC--1991-03-01 UTC
5 2001-03-01 2001-11-01 2001-03-01 UTC--2001-11-01 UTC
6 2007-12-01 2009-06-01 2007-12-01 UTC--2009-06-01 UTC
# Get every day in the range of dates
my_dates <- seq(first(my_df$recession_start), today(), by = "day")
# Create a list of intervals
recession_intervals <- list(my_df$recession_interval)
# Check to see if `my_dates` are in the intervals
recession <- my_dates %within% recession_intervals # Throws warning and does not give expected results
I suspect this is because my list of dates is a single list vs. multiple lists as in the tidyverse example, but I'm not sure how to create multiple lists other than manually.
The desired output would be a data frame with each date and a "TRUE" or "FALSE" column indicating if that daily date is in a recession interval. Something like:
recession_df <- data.frame(Date = my_dates, recession = recession)
Output would look like this:
Date recession
1 1973-11-01 TRUE
2 1973-11-02 TRUE
3 1973-11-03 TRUE
4 1973-11-04 TRUE
5 1973-11-05 TRUE
6 1973-11-06 TRUE
7 1973-11-07 TRUE
8 1973-11-08 TRUE
9 1973-11-09 TRUE
10 1973-11-10 TRUE
Thanks for any help!
One option is to loop over (map) the 'my_dates', check if there are any dates that are %within% the 'recession_interval column, create a tibble with each 'date' and the logical output and convert to a single dataset with _dfr (row binding)
library(purrr)
out <- map_dfr(my_dates, ~ tibble(Date = .x,
recession = any(Date %within% my_df$recession_interval)))
-output
# A tibble: 17,381 x 2
Date recession
<date> <lgl>
1 1973-11-01 TRUE
2 1973-11-02 TRUE
3 1973-11-03 TRUE
4 1973-11-04 TRUE
5 1973-11-05 TRUE
6 1973-11-06 TRUE
7 1973-11-07 TRUE
8 1973-11-08 TRUE
9 1973-11-09 TRUE
10 1973-11-10 TRUE
# … with 17,371 more rows