Feature proposal/idea: clean.conc.na using CTROUGH

[Gero1999]

@billdenney Hey, if I am right is a common standard for steady state studies to use "ctrough" as an imputation method for missing data in NCA. The only challenge I can see would be considering the end argument in the function when calling pk.calc.ctorugh.

Do you think would be feasible to add this additional option in PKNCA for clean.conc.na?

[billdenney]

This would not be a clean.conc.na() change but an imputation method change.

Most often, I see cmin used rather than ctrough, but I see both as useful.

We currently have PKNCA_impute_method_start_cmin(), and adding PKNCA_impute_method_start_ctrough() would be a good addition. We could also add PKNCA_impute_method_end_cmin() and PKNCA_impute_method_end_ctrough() for completeness.

[Gero1999]

That is indeed very clever, PKNCA_impute_method_end_cmin & PKNCA_impute_method_end_ctrough will solve this problem.

Regarding how these functions should behave in "weird" scenarios; all time NA, all conc NA (#361), I would set it here as:

1. All concentrations NA >> Returns NA
2. All times are NA >> Error (Should not indeed happen within the PKNCA workflow, right?).