Reprex below. <div class="highlight highlight-source-r notranslate position-relati

Date-type column causing select.() to fail about tidytable HOT 2 CLOSED

markfairbanks commented on July 19, 2024

Date-type column causing select.() to fail

from tidytable.

Comments (2)

markfairbanks commented on July 19, 2024 1

If you update to the latest CRAN version (v0.4.1) this problem won't occur anymore.

You can also use any predicate you want now. In the example below you'll see I use lubridate::is.POSIXct.

library(tidytable, warn.conflicts = FALSE)
library(lubridate, warn.conflicts = FALSE)

df1 <- structure(list(
    id = "cal", sequence_id = "0357C554_DA8B_4D2A_945B_8381F2CE0C48",
    date = structure(1548028800, class = c("POSIXct", "POSIXt"), tzone = "UTC")
  ), row.names = c(NA, -1L), class = "data.frame")

df1 %>%
  select.(id)
#>     id
#> 1: cal

df1 %>% 
  select.(id, lubridate::is.POSIXct)
#>     id       date
#> 1: cal 2019-01-21

It's worth noting that in v0.4.1 there's a pretty big change that might affect you since you've been using tidytable for a while. You can no longer use by = list(col1, col2), you have to use by = c(col1, col2).

test_df <- data.table(a = c("x","x","y"),
                      b = c("x","x","y"),
                      c = 1:3)

# Using by = list(a, b) no longer works
test_df %>%
  summarize.(avg_c = mean(c), by = list(a, b))
#> Error: Using by = list(col1, col2) is deprecated. Please use by = c(col1, col2)

# Use by = c(a, b)
test_df %>%
  summarize.(avg_c = mean(c), by = c(a, b))
#>    a b avg_c
#> 1: x x   1.5
#> 2: y y   3.0