One popular approach is pivot_longer
from tidyr
.
library(tidyr)
data %>%
pivot_longer(-created_at,names_to = "x",values_to = "y")
# A tibble: 4 x 3
# created_at x y
#*
#1 2020-03-06 negative 0.890
#2 2020-03-06 positive 0.756
#3 2020-03-07 negative 0.904
#4 2020-03-07 positive 0.799