Stop_Num

[IJBG]

To get a good level of detail of landscape change, planning on splitting the routes into half or quarter routes for analysis. To do this, need to:

• [-] Find a way to extract stop-num that may be less fragile (import_ebird_data.R) and therefore captures the missing extracts from ebird
• [-] Function to add stop_num = 1 to routes in a year that have 20 stops and every other stop_num BUT stop 1
• [x] Look at Haven's papers + old website files to figure out which routes are going to have the stop-level data going all the way back.
• [-] Add stop_num in for the prev. website data we've got it for
• [-] Some ebird data ie: https://ebird.org/checklist/S18551742 does have the stop information in the notes from each bird. Think about how to tackle this - is there a good way to extract this data from the ebird imports? Doing this manual would be quite bad, there 50's checklists like this. List of these routes in scratch file (once finished add scratch files back to .gitignore (scratch**)
  • [ ] ,,,,,,1,,,,
  • [ ] in overall checklist note by quarter stops
  • [ ] Stops: 0,1,0
  • [ ] Stops = 0,1,0,
  • [ ] 1=0, 2=15,3=
  • [ ] stop 5 =1, stop 9 = 18,
  • [ ] st 6 OR st14=2; st18=1;
• [x] Add species_comments column to import_ebird_data so that the above data ^ is in the mbbs and can be worked with
• [x] re-process this year's data to include the new species_comments column + this will fix Rokoske errors
• [-] Add date and time to mbbs_survey_events
• [x] Update the processing for identifying the main observer. Based on mbbs meeting and talking with folks, separate the first three observers out into separate columns [this will go into the survey events df]. Give unique ID for combos of observers. Alternative would be assigning primary observer to whomever of the first three observers has the highest rank. Exceptions for Chatham 2 and 4, Norm and Mike's routes will stay listed under Norm and Mike. Also exception for Jeff Pippen route?
• [x] Added an observer_rank as a fixed effect, based off modeling notes
• [-] convert generate observer_ID into a function that takes the desired column and generates an ID from that

Along the way:

• [x] Add fixes to two common_name errors from the old website data (GHOW and House Wren)
• [x] Add a function to the package to merge the mbbs_county datasets - all the ease of functionality of having the merged dataset, none of the extraneous data storage. Maybe this goes in a new script in R/ ?
• [-] #2
• [-] Update packages associated with mbbs in the... lockfile?
• [x] Quality of life so may not be necessary to update until next year, BUT - the function update_observer_table in R/process_observers should check if the mbbs_county/rt/year for a given route/observer combo is already on the survey_list, and if so, do not prompt for what the new observer should be if is.na(observers == TRUE). If the data on the observer has already been fixed in the survey_list and therefore in data/mbbs_survey_events, there's not actually an NA there. That's part of the whole point of mbbs_survey_events. Having to confirm that there are some rows in the dataset for which observers= has failed to propagate every year is not necessary.
• [-] Move count to earlier in the mbbs_ files, and put count_raw like at the end. High potential for confusion for new users

[IJBG]

Left to do from today to finish observers section

• [x] combine obs1 obs2 obs3 into standardized_observers (with consistent ordering)
• [x] integrate scratch_new_observer_processing back into process_observers
• [x] decide on renaming current observers column to raw_observers, so then 'final observers' is just the observers column, additionally where to do that to minimize code disruption. Or just replace observers in final dataset with final observers then select(-finalobservers) <<probably this last one - "raw observers" and "observers_standardized"
• [x] QoL update described in main post above, stop from prompting for NAs when that data exists on survey_events
• [x] turn QoL update into a function

[IJBG]

@bsaul If the last remaining check() warning is this, should I run the suggested R CMD build --resave-data?

❯ checking data for ASCII and uncompressed saves ... WARNING

Note: significantly better compression could be obtained
      by using R CMD build --resave-data
                       old_size new_size compress
mbbs.rda                  955Kb    452Kb    bzip2
mbbs_chatham.rda          476Kb    207Kb    bzip2
mbbs_durham.rda           185Kb    106Kb       xz
mbbs_orange.rda           305Kb    152Kb    bzip2
mbbs_survey_events.rda     10Kb      9Kb    bzip2