- added new functionality for
slice_*, as in dplyr, including head, tail, max, min and sample. These include the dplyr arguments such asn,prop, andweight_by(for sampling). - added
join_mcols_left(x,y)for GRangesxand tabulary - added
distance=TRUEoption forjoin_overlap_*
- added
distance=TRUEoption forjoin_overlap_* - added
join_mcols_left(x,y)for GRangesxand tabulary
- added
pull()method for Ranges objects that extracts a single column as a vector - moved
dplyrto Depends and removing tidyverse reexports
- minor spelling and layout fixes to vignette,
- @PeteHaitch corrected table layout in vignette
- minor documentation fixes
-
Spencer Nystrom has made several significant contributions to the
join_nearestfamily of functions:join_nearest_(x, y, ..., distance = TRUE)family of functions now takes a new argument,distance, which allows the user to add a column for the distance of the nearest rangeyto that inx.add_nearest_distance_(x, y, ...)family of functions, which will add a new metadata column to thexranges object which contains the distance to its nearest neighbor iny. If there are no nearest neighbors, the new column will be given a missing value.
- refactoring of select internals, improved speed when casting a GRanges -> DataFrame
- further fixes to reduce/disjoin internals
- fixes reduce/disjoin internals cleans up disjoin cases when an expansion occurs
- set tidyselect version to be v 1.0
- set coverage method for delegating ranges
- fix docs for bam reading
- move from
tidyselect::vars_select()totidyselect::eval_select()
- update handling of list columns,
expand_ranges()no longer takes cartesian product if lists are parallel.summarize()properly handles list column output without blowing out number of columns.
- adds method for
dplyr::sample_n()
- fixed issue #62 for Ranges construction, the
as_granges()andas_iranges()functions now handle List columns correctly - added in helper functions for dealing with names in Ranges.
See
?ranges-namesfor details.
- added
slice()for Ranges, and GroupedRanges - internals of grouping have been overhauled, but there shouldn't be any user facing changes. It is now much faster to generate groupings.
- group information can be interrogated with
dplyr::group_keys() - a GRangesList can be obtained automatically from a GroupedGenomicRanges with
dplyr::group_split() - group indices can be generated with
dplyr::group_indices()
shift_downstream()andshift_upstream()now properly handle vector amounts ofshift. Fixes issue #73
- Left outer join overlap operations now work if either
xoryhave no metadata columns see #70 - Left outer join overlap operations will also correctly behave in situations when there are no non-overlapping ranges.
- Left outer join overlaps no longer modify seqinfo (see here)[https://support.bioconductor.org/p/125623/]
- patch left outer join when
xoryare IRanges, flesh out overlaps documentation.
- Reformatting
NEWS.mdso no longer softlinks to inst/NEWS
- plyranges release and devel have removed
unnest()and replaced it withexpand_ranges()due to changes in the tidyr API. Please replace all uses of this function withexpand_ranges()
- fixed bind_ranges so it preserves rownames
- enable right generics to be called upon invoking plyranges functions without loading plyranges
- added tile/window methods
- fixed up documentation
- doc updates
- speed up of
group_bymethods - refactor of BAM reading utilities
- refactored
set_widthout so it's called internally by mutate - along with
set_widththere are other internalset_methods - add
_within_directedvariants for overlaps methods - modified
overscope_rangesto be an S3 method, should enable more refactoring in the future
https://bioconductor.org/packages/devel/bioc/html/plyranges.html
- package passed review and is now on Bioconductor devel branch!
- I've been pretty slack with updating the NEWS file but will be more diligent in the future.