Minor fixes to fake data

[Aastedet]

Small edits to provide:
a) The correct NPU code length (8)
b) to_wwyy() a way to simulate edge cases of events that fall on isoweek 53 of a year (I think lubridate might already have handled these correctly on linux, but not on windows)
c) LPR3's diagnoser$diagnosetype now match LPR2's lpr_diag$c_diagtype (I just did a simple simulation_definitions[25,]$generator <- simulation_definitions[16,]$generator)

[Aastedet]

Small edits to provide: a) The correct NPU code length (8) b) to_wwyy() a way to simulate edge cases of events that fall on isoweek 53 of a year (I think lubridate might already have handled these correctly on linux, but not on windows) c) LPR3's diagnoser$diagnosetype now match LPR2's lpr_diag$c_diagtype (I just did a simple simulation_definitions[25,]$generator <- simulation_definitions[16,]$generator)

And I just realized c) is already taken care of in #86 (that's what I get for not reviewing PR's 😜 )