library(tidyr)

df <- tribble(
  ~Date, ~Fruit, ~Sold,
  "2020-02-01", "Apple", 5,
  "2020-02-01", "Banana", 1,
  "2020-02-02", "Apple", 2
)

df_complete <- df %>%
  complete(Date, Fruit)

df_complete
#> # A tibble: 4 x 3
#>   Date       Fruit   Sold
#>   <chr>      <chr>  <dbl>
#> 1 2020-02-01 Apple      5
#> 2 2020-02-01 Banana     1
#> 3 2020-02-02 Apple      2
#> 4 2020-02-02 Banana    NA

df_complete[is.na(df_complete$Sold), "Sold"] <- 0

df_complete
#> # A tibble: 4 x 3
#>   Date       Fruit   Sold
#>   <chr>      <chr>  <dbl>
#> 1 2020-02-01 Apple      5
#> 2 2020-02-01 Banana     1
#> 3 2020-02-02 Apple      2
#> 4 2020-02-02 Banana     0

# What if there are more columns to complete? Gets cumbersome.

df_more <- tribble(
  ~Date, ~Fruit, ~Sold, ~Ate, ~Garbage,
  "2020-02-01", "Apple", 5, 1, 4,
  "2020-02-01", "Banana", 1, 2, 1, 
  "2020-02-02", "Apple", 2, 1, 1
)

df_more_complete <- df_more %>%
  complete(Date, Fruit)

df_more_complete[is.na(df_more_complete$Sold), "Sold"] <- 0
df_more_complete[is.na(df_more_complete$Ate), "Ate"] <- 0
df_more_complete[is.na(df_more_complete$Garbage), "Garbage"] <- 0

df_more_complete
#> # A tibble: 4 x 5
#>   Date       Fruit   Sold   Ate Garbage
#>   <chr>      <chr>  <dbl> <dbl>   <dbl>
#> 1 2020-02-01 Apple      5     1       4
#> 2 2020-02-01 Banana     1     2       1
#> 3 2020-02-02 Apple      2     1       1
#> 4 2020-02-02 Banana     0     0       0

# Or you could just do that all within complete()

df_more_complete <- df_more %>%
  complete(Date, Fruit, fill = list(Sold = 0, Ate = 0, Garbage = 0))

df_more_complete
#> # A tibble: 4 x 5
#>   Date       Fruit   Sold   Ate Garbage
#>   <chr>      <chr>  <dbl> <dbl>   <dbl>
#> 1 2020-02-01 Apple      5     1       4
#> 2 2020-02-01 Banana     1     2       1
#> 3 2020-02-02 Apple      2     1       1
#> 4 2020-02-02 Banana     0     0       0

# What if you want to do a pipeline?
library(dplyr, warn.conflicts = FALSE)

df %>%
  complete(Date, Fruit, fill = list(Sold = 0)) %>%
  group_by(Date) %>%
  summarise(Total = sum(Sold))
#> # A tibble: 2 x 2
#>   Date       Total
#>   <chr>      <dbl>
#> 1 2020-02-01     6
#> 2 2020-02-02     2
           
# If you do is.na() etc, have to break it:

df_na_filled <- df %>%
  complete(Date, Fruit)

df_na_filled[is.na(df_na_filled$Sold), "Sold"] <- 0

df_na_filled %>%
  group_by(Date) %>%
  summarise(Total = sum(Sold))
#> # A tibble: 2 x 2
#>   Date       Total
#>   <chr>      <dbl>
#> 1 2020-02-01     6
#> 2 2020-02-02     2