You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Fix: in readLAScatalog() the documentation states that ... is propagated to list.files(), but the argument pattern was actually hard coded internally and this prevents it being overwritten. When using readLAScatalog(..., pattern = "xxx") this previously triggered an error, formal argument "pattern" matched by multiple actual arguments. It now works. See #368.
Fix: in spTransform() the reprojected point cloud now has quantized coordinates and is thus LAS compliant #369.
Fix: The local maximum filter algorithm more robustly finds local maxima when two or more close points or pixels share the exact same height and are both locally the highest. Previously, if two or more points in a close neighbourhood were both the highest, they may all be missed depending on the order they were processed (which is somewhat random). The fix allows users to retain one local maximum among multiple ones with a precedence to the first one identified as local maximum. The consequences of this fix are that slightly more apices may be found, especially when processing a CHM in RasterLayer.
Fix: classify_ground() no longer erases the previous classification when no ground points were recorded but some points are classified with other classes.
Fix #365. Poor interpolation at the very edge of the Delaunay triangulation in some cases. Triangles with too steep a slope are now removed. This triggers a knnidw interpolation instead.
Fix #371: las_reoffset() may not have caught extremely rare Z coordinate overflow when converting to integers.
Fix #372. las_reoffset() incorrectly converted decimal coordinates to integers using trunc instead of round.
Fix: projection<-() and crs<-() properly attributes NA CRS for LAS 1.4 objects
Change: in print the CRS of LAS and LAScatalog is no longer displayed as a proj4 string but uses the WTK string with sf style display. E.g. NAD83 / UTM zone 17N is displayed instead of +proj=utm +zone=17 +datum=NAD83 +units=m +no_defs. This is part of the migration toward WTK instead of proj4.
Change: lidR now explicitly depends on rgdal >= 1.5.8.
Change: grid_canopy() now rounds the values of the pixels for not outputing pixels that with an irrelevant number of decimal digits.
Enhance: epsg() now throws a warning if the LAS is in format 1.4 and CRS is stored as WKT.
New: projection()<- supports crs from sf and numeric values for espg code: projection(las) <- 26918.
New: in spTransform() it is now possible to use a parameter scale to change the scale factor after reprojection. This is useful for projecting from lon-lat data las2 = spTransform(las, crs, scale = 0.01).
Internal: better support in projection<- of the current changes with CRS representation in the R spatial ecosystem.