calibrateFireRegimePolys(): when no spread probability in the calibration range gives a
polygon's mean fire size, it now uses the end of the range whose fitted size is closer, not always
the bottom of the range. A polygon whose mean fire size was just above the range got the lowest
spread probability, so its fires barely spread;comparePredictions_summaryDT(): the achieved fire return interval (and the internal mean annual area
burned) again count the area burned inside the reporting area, from fires ignited there or outside
(burnSummary groups 1 and 3). This fixes a regression in scfmutils 0.0.9 (cbcd0b5): the
groups-1-and-3 table was overwritten by the groups-1-and-2 one used for mean fire size, so a fire
that crossed the reporting boundary was counted by where it was ignited, not where it burned;calcIgnitions(), countBurnedPixelsInSAR(), and fireRegimeParRaster(),
moved from the scfmIgnition, scfmSpread, and scfmEscape/scfmIgnition/scfmSpread modules
(#13). calcIgnitions() drops the module version's unused fireRegimePolys argument, and
countBurnedPixelsInSAR() takes the reporting-area pixels as pixInSAR instead of reading them
from the module;calibrateFireRegimePolys() no longer fails ('xmin' not less than 'xmax') when every flammable
cell of a fire regime polygon has 8 flammable neighbours, e.g., with no non-flammable cells
(PredictiveEcology/scfm#22): p0 is then hatP0(pEscape, 8) exactly;calibrateFireRegimePolys() estimates p0 correctly from 4-neighbour counts: escapeProbDelta()
assumed 0 to 8 neighbours, so with neighbours = 4 it recycled the weights and underestimated p0;getFirePoints_NFDB_scfm() downloads from .../current_version/NFDB_point_shp.zip: CFS renamed the
archive from NFDB_point.zip, which now returns HTTP 404, so no newer release could be fetched;withr to Imports from Suggests, for use in examples;prepInputsFireRegimePolys() for BEC;prepInputsFireRegimePolys() gets a new argument subsetType to retrieve entire polygons
bordering studyArea or only the spatial intersection (default, preserving previous behaviour)scfmDiagnostics plots.makeLandscapeAttr() properly handles cases with missing neighbours;calibrateFireRegimePolys() gracefully exits if polygons have no fires;withr to Suggests (used with testing);reproducible >= 2.1.0;reproducible: pass to arguments to reproducible::prepInputs() and related functions, instead of passing studyArea or rasterToMatch;scfmutils.driver.plot.scam (default TRUE) to allow user to disable creation of scam plots in calibrateFireRegimePolys();calibrateFireRegimePolys();fpCompare to Imports, for improved floating point number comparisons;SpaDES.core, SpaDES.tools, and reproducible;.makeLandscapeAttr() that was introduced by switching to sf for fire regime polygons;comparePredictions_fireDistribution() to use arguments fireRegimePoints, burnSummary, and size instead of dt;checkForIssues();sf polygons object instead of lists of polygon attributes;Require, reproducible, SpaDES.core, SpaDES.tools;ggplot wrappers for plotting scfm rasters;Transitional version: introduces several breaking changes
sf polygons object instead of lists of polygon attributes;cellSize and neighbours from calibrateFireRegimePolys();calcZonalRegimePars() now returns a data.table instead of a list;prepInputsFireRegimePolys() when url supplied by user;LandR dependency;bcdata to Suggests to allow use of BEC zones etc. prepInputsfireRegimePolys();makeLandscapeAttr();googledrive to Suggests;SpaDES.core and SpaDES.tools versions following terra migration and other changes;prepInputsFireRegimePolys() example to use terra;terra instead of raster package;fasterize dependency;magrittr's;studyAreaReporting;%>% for pipe (R < v4.2);prepInputsFireRegimePolys();prepInputsFireRegimePolys() from BEC NDTs: allow either col name;plotPath exists for scam plots;comparePredictions_summaryDT();deSliver() and clarify documentation;scam() in calibrateFireRegimePolys();scfmRegime functions;checkForIssues() and deSliver() for working with fire regime polygons;prepInputsFireRegimePolys();SpaDES.core version requirement;PolyId to PolyID;executeDesign();theme_bw() instead of theme_minimal() in diagnostic plots;fireRegimePolyTypes();comparePredictions_annualIgnitions();times argument to comparePredictions_fireReturnInterval();simList to comparePredictions_summaryDT();prepInputsFireRegimePolys();startTime to .executeDesignInternal();.makeLandscapeAttr();scfmLandCoverInit;inherits;prepInputsFireRegimePolys();pkgdown for website;