Cabinet, election and party information can now come from REPDEM (new data releases, class voteListRepdem) or ParlGov (older releases). All functions work with both layouts.
addCabinetInfo() uses the new cabinetPeriods and voteCabinet tables for REPDEM data. caretaker is then 1 for every caretaker (demissionair) period: after resignation, after an election, and for interim cabinets. REPDEM data also get cabinet_id, cabinet_period_id, caretaker_reason and caretaker_parlgov. For ParlGov data, caretaker is taken from cabinetInfo as before.
addPartyInfo() matches cabinet parties on cabinet_id when available, ignores duplicate rows in partyElectionInfo, partyCabinetInfo and partyInfo, and sets cabinet_party and prime_minister to 0 for parties that hold seats but are not listed for the cabinet in office. REPDEM data also get support_party; includetype = "all" adds family_name, ches_lrgen, ches_lrecon and rile. Cabinet participation follows the same cabinet as addCabinetInfo(): voteCabinet/questionCabinet for votes and responders (date as fallback), the question date for questioners.
addCabinetInfo() for questionList objects no longer fails when cabinetInfo has no formal_resignation column.
filter() and subset() keep the voteCabinet, questionCabinet and questionText tables in sync. filter() on a voteList can filter on .table = "voteCabinet".
addCabinetInfo() for questionList objects uses the questionCabinet table of REPDEM data (cabinet at the date of the response).
examplevotes and examplequestions now use the REPDEM layout. The same votes and questions with ParlGov data are available as examplevotes_legacy and examplequestions_legacy. Text in the example data is ASCII only.
dutchparl 0.0.0.9009
addPartyInfo() and rice() no longer warn about unknown columns when the data are stored as tibbles.