Numbers are rounded by three mechanisms:
Each column may have an attribute "round" with the numbers of digits. This results in a character vector with a fixed number of digits. NA values are replaced by an empty string.
The data frame may have a column ".digits" with a number of digits. All numeric values are rounded to the number of digits in this column, the column is discarded afterwards.
The data frame may have an attribute "digits" resulting in all numbers in the data frame being rounded to that number. The constant VLKR_NORMAL_DIGITS is used as a fallback.
.knit_table(df, ...)
Formatted table produced by kable.
Data frame.