Codebookr - L2C Master Log

[areddy647712]

Hi Brad,

This is the error I received when passing the Master Log data frame to the codebook function.

Error in dplyr::mutate():
! Problem while computing Value = date.
✖ Value must be a vector, not a function.

These are the steps I took:

1. Load Master_Log version2
2. glimpse(Master_Log version2)
3. Pass the data frame to the codebook function

I hope this was helpful. I also tried to use the cb_add_col_attributes() function and assign attributes to columns with dates in them.

Thank you!
Aish

[mbcann01]

Thank you!

[mbcann01]

POSIXct issue

It looks like this error is caused by POSIXct columns. I'm not sure why yet. It also looks like the code that is struggling is on lines 72-76 of cb_summary_stats_time.R

It doesn't look like the problem is with a POSIXct column per se. The date_time column I created in the example study data works just fine. Not, sure what the problem is.

The problem is with calculating the mode when there is a mode value (as opposed to all values occurring exactly one time).

The section of code that calculates the mode value still had the date column hardcoded into it from when I originally created and test the function. This explains why the date column had a mode value and didn't throw an error, but other columns with mode value did.

mode <- df %>%
      dplyr::count(.data[[.x]], name = "Frequency") %>%
      dplyr::filter(Frequency == max(Frequency)) %>%
      dplyr::mutate(
        Statistic = "Mode",
        Percentage = Frequency / nrow(df) * 100,
        Value = date
      )

Changing the last line Value = date to Value = .data[[.x]] fixes the problem.

• Update the study data to include a POSIXct column
• Fix the problem
• Unit test (decided I don't really need this. The code for the new test wasn't any different than the code for the previous test. It was just a different column name).
• Update documentation
• Check
• Merge to main
• Install
• Push