TODO.md

THOUGHTS

• My intention with the output is that the objects are simple enough for users to interact with directly ...
• Implement diagnostics, assessment and exploration of the input data.
• Implement diagnostics, assessment and exploration of the output result objects.
• Could the review of these data objects be streamlined using classes and methods?
• If so, how should these be structured?

TODO

Phase out plyr functions.

The make_whole() function

• provide helpful output (verbose = TRUE), such as
  • the start date,
  • the end date,
  • total number of NAs,
  • total % NAs,
  • mean % NAs per year

The detect() function

• produce helpful output (verbose = TRUE), such as
  • the start date of the time series,
  • its end date,
  • the climatological mean temperature,
  • the climatological coldest temperature,
  • the climatological warmest temperature,
  • the threshold mean,
  • the threshold min,
  • the threshold max,
  • the number of events of <= 5 days in duration,
  • the number of events of 5 <= 10 days in duration,
  • the number of events of 10 <= 20 days in duration,
  • the metrics top three longest events, i.e.
    • date_start,
    • date_peak,
    • date_stop,
    • duration,
    • int_mean,
    • int_max,
    • int_cum,
    • int_var,
    • rate_onset,
    • rate_decline
  • w.r.t. the above, another option might be to have a summary() function act on the detect() function's output; probably the first choice