Convert character to date vector (!bY) in R

Question

I am trying to convert a character vector of dates (in the format: i.e. "Jan.1990") to a date vector (keeping a similar format: i.e. "Jan 1990" or "Jan.1990").

month_year <- ("Jan.1990", "Feb.1990", "Mar.1990", "Jan.1991", "Feb.1991", Mar. 1991")

I have tried using as.Date and various commands with the lubridate package but I either end up with an incorrect format or NAs.

I tried using as.Date:

df$month_year <- as.Date(df$month_year,format = "%b%Y")

This resulted in NAs.

I tried lubridate::parse_date_time2

parse_date_time2(tidy_labor$month_year, c("Jan.1990"), exact = TRUE, orders = "bY")

But the format came out as:

unknown timezone 'Jan.1990'unknown timezone 'Jan.1990'   [1] "1990-01-01 GMT" "1990-01-01 GMT" "1990-01-01 GMT" "1990-01-01 GMT" "1990-02-01 GMT" "1990-02-01 GMT" "1990-02-01 GMT"

Any help would be much appreciated.

A month and a year is not a "date" as you need a day for that. But there are `yearmon` types... — Dirk Eddelbuettel, Apr 04 '17 at 03:08

score 0 · Answer 1 · answered Apr 04 '17 at 03:13

0

This question comes up a lot and the short answer is that dates require a day.

You could use zoo::as.yearmon:

library(zoo)
as.yearmon("Jan.1990", "%b.%Y")

Or you could assume that the day is always the first of the month, concatenate that to your values and convert to Date type.

answered Apr 04 '17 at 03:13

neilfws

32,751
5
50
63

1

If it comes up a lot we should vote to close it as a duplicate. – Rich Scriven Apr 04 '17 at 03:15
Yes, there are so many minor variants of it, I can't be bothered searching for the best duplicate :) so it's good that you can. – neilfws Apr 04 '17 at 03:19

Convert character to date vector (!bY) in R

1 Answers1