NEWS.md
fracture() and frac_mat() no longer return an error when receiving a vector containing NA or Inf values (#14).fracture(NA) now returns NA.fracture(Inf) now returns "Inf/1".frac_style() would print integers in mixed fractures as "1 ⁰/₀" instead of "1" (#12).fracture() and frac_mat() is now ..., which must be empty. As a result, all arguments besides x must now be named. (#5)fracture() and frac_mat() now default to a denominator of 1 when x equals 0 or 1. Previously, these would default to a denominator of max_denom. max_denom is still used as the denominator when x is 0 or 1 ± ε. (#6)fracture() and frac_mat() gain the argument denom, which allows the user to set an explicit denominator used by all fractions. (#5)frac_mat(1307.36, base_10 = TRUE)) would cause an floating point rounding error.decimal_to_fraction_base_10() with double rather than int.frac_mat()’s max_denom should be less than the inverse square root of machine double epsilon.frac_style() which formats fractures using Unicode superscripts and subscripts.is.frac_mat() which tests if a matrix is formatted like the output of frac_mat().decimal_to_fraction_cont() with double rather than int.NA and infinite inputs to frac_mat().