Handling missing values with tidyr::fill()

[ghost]

If missing value appears first in its key and it needed to be replaced by last observed value, it doesn't leave it with NA and cares last observed value from previous key (group).

harvest <- tsibble(
  year = c(2011, 2013, 2014, 2010, 2012, 2014),
  fruit = rep(c("kiwi", "cherry"), each = 3),
  kilo = sample(1:10, size = 6),
  key = id(fruit), index = year
)

harvest_fill <- fill_na(harvest, .full = TRUE)

harvest_fill %>%
  group_by_key() %>% # I've tried also group_by
  tidyr::fill(kilo, .direction = "down")

2010 kiwi observation need to be leaved with NA. I accomplished it by:

harvest_correct_fill <- 
  harvest_fill %>% 
    as.data.frame() %>% 
    group_by(fruit) %>% 
    tidyr::fill(kilo, .direction = "down")

[earowang]

Thanks. I've added fill.grouped_ts to fix the issue. And fill() is re-exported from tidyr, so we can remove tidyr::.

library(tsibble)
harvest <- tsibble(
  year = c(2011, 2013, 2014, 2010, 2012, 2014),
  fruit = rep(c("kiwi", "cherry"), each = 3),
  kilo = sample(1:10, size = 6),
  key = id(fruit), index = year
)

harvest_fill <- fill_gaps(harvest, .full = TRUE)
harvest_fill %>%
  group_by_key() %>%
  fill(kilo, .direction = "down")
#> # A tsibble: 10 x 3 [1Y]
#> # Key:       fruit [2]
#> # Groups:    fruit [2]
#>     year fruit   kilo
#>    <dbl> <chr>  <int>
#>  1  2010 cherry     9
#>  2  2011 cherry     9
#>  3  2012 cherry     6
#>  4  2013 cherry     6
#>  5  2014 cherry     2
#>  6  2010 kiwi      NA
#>  7  2011 kiwi       1
#>  8  2012 kiwi       1
#>  9  2013 kiwi       4
#> 10  2014 kiwi       7

^{Created on 2018-11-21 by the reprex package (v0.2.1)}