Take times without a unit to be hours when detecting the dosing interval

[billdenney]

Summary

• Without a time unit, find.tau(), find.dose.regimen(), choose.auc.intervals(), PKNCAdata() and pk.nca() now match the dosing interval to the built-in nominal intervals in hours, the same assumption PKNCA already makes for date-times. find.tau(c(0, 24, 50)) is 24 (once daily) rather than the median spacing of 25.
• A stated timeu overrides the assumption. A stated unit that cannot be converted to hours still leaves the interval to the dose times. timeu = NA says so explicitly.

Behaviour changes

• Unitless dose times recorded a little early or late give the nominal interval (24 rather than 24.5 h). That moves the interval generated for the last dose and multiple-dose parameters such as mrt.md.obs.
• Unitless data whose interval matches no nominal interval in hours, such as data recorded in minutes or days without saying so, or dosing every hour, now give pknca_warning_tau_not_nominal. State timeu in PKNCAconc() to avoid it.

Tests

• Tests that pinned data-derived medians without a unit now expect the snapped values, with the data-derived values pinned under timeu = NA. Tests of times in days, minutes or abstract units assert the new warning or pass timeu = NA.
• The full suite passes, and R CMD check gives 0 errors and 0 warnings.

🤖 Generated with Claude Code