How do I initialize an empty data frame *with a Date column* in R?

user_12135864 0
Viewed 772

How do I initialize an empty data frame with a Date column in R?

I've tried a number of things, but they all generate an error message when I try to append a row with a date in the correct column. For example:

> test <- data.frame(date=as.Date(character()), var1 = as.numeric(), var2 = as.numeric())
> str(test)
'data.frame':   0 obs. of  3 variables:
 $ date: 'Date' num(0) 
 $ var1: num 
 $ var2: num 
> test <- rbind(test, c(as.Date("2020-01-01"), 123, 456))
Error in as.Date.numeric(e) : 'origin' must be supplied

Initializing the first column as an integer and then adding the "Date" class to it generates the same error:

> test <- data.frame(date=integer(0), var1=as.numeric(), var2=as.numeric())
> str(test)
'data.frame':   0 obs. of  3 variables:
 $ date: int 
 $ var1: num 
 $ var2: num 
> class(test$date) <- "Date"
> str(test)
'data.frame':   0 obs. of  3 variables:
 $ date: 'Date' int(0) 
 $ var1: num 
 $ var2: num 
> test <- rbind(test, c(as.Date("2020-01-01"), 123, 456))
Error in as.Date.numeric(e) : 'origin' must be supplied

I've also searched for this error message, but the examples I have found all deal with converting a number into a date. I haven't been able to find any examples in my context.

What's the correct way to initialize the date column so that I can add formatted dates to it as above?

Thanks!

dataframe r date
2 Answers

The issue is more likely due to concatenation as vector cannot have multiple class. We can pass a data.frame

rbind(test, setNames(data.frame(as.Date("2020-01-01"), 123, 456), 
        names(test)))

-output

        date var1 var2
1 2020-01-01  123  456
user_37322710 
library(tidyverse)
df = tibble(
  date = as.Date(x = integer(0), origin = "2000-01-01")
)
user_166717360
Related