Fill missing values with complete using dplyr-CodePudding

I have a dataframe like this:

library(tidyverse)

df = tibble(AOI = c('NA', 'NA', 'NA', 'NA', 'NA', 'EA', 'EA'),
            Values = c(-1, -1, 0, 1, 1, -1, 0))

df = df %>% group_by(AOI, Values) %>% tally()

AOI     Values  n
<chr>   <dbl>   <int>
EA     -1       1
EA      0       1
NA     -1       2
NA      0       1
NA      1       2

And I would like to ensure that all the values -1, 0 and 1 exist for each unique AOI. As of right now the group EA is missing 1. I would like to use I believe complete to ensure any of the three values that are missing are automatically added and n is filled with 0.

I am trying this:

df = df %>% complete(AOI, Values, fill = list(n = 0))

but this fails with:

Error in `dplyr::summarise()`:
! Problem while computing `..1 = complete(data = dplyr::cur_data(), ...,
  fill = fill, explicit = explicit)`.
ℹ The error occurred in group 1: AOI = "EA".
Caused by error:
! object 'AOI' not found

My expected output is this:

AOI     Values  n
<chr>   <dbl>   <int>
EA     -1       1
EA      0       1
EA      1       0
NA     -1       2
NA      0       1
NA      1       2

EDIT:

It doesn't have to be complete to get this task done, any way to get the desired output is fine with me.

CodePudding user response：

I assume after constantly overwriting your dataframe a few times, it may have produced an error, maybe for saving something unexpected.

*So, I changed the code a little.

Code

library(tidyverse)

df <- tibble(AOI = c('NA', 'NA', 'NA', 'NA', 'NA', 'EA', 'EA'),
            Values = c(-1, -1, 0, 1, 1, -1, 0))

df %>%
  arrange(AOI) %>% 
  count(AOI, Values) %>% 
  complete(AOI, Values,fill = list(n = 0))

Output

# A tibble: 6 x 3
  AOI   Values     n
  <chr>  <dbl> <dbl>
1 EA        -1     1
2 EA         0     1
3 EA         1     0
4 NA        -1     2
5 NA         0     1
6 NA         1     2