Minor changes:
Minor changes:
unlist2
: Added support for NULL
definition in nameCol
and rownameCol
variables.New functions:
filterNested
: Keep rows that match a nested condition. Performs recursive
regular expression matching across all columns, including nested list columns.New functions:
unlist2
: An alternative method to unlist
for DFrameList
that does a
better job of keeping track of names
and rownames
in the original object,
which makes conversion into long format simpler.Minor changes:
"by"
argument. Improved error messages on type mismatches and non-atomic input.Major changes:
Minor changes:
rbindToDataFrame
: Added an additional assert check.Minor changes:
leftJoin
and rightJoin
now allow for NA
values defined in the by
column for the x
side of the join. This was required for OncoTree metadata
additions in the pending Cellosaurus package update.by
columns are atomic.unnest2
now uses parallell mclapply
for the main unnest step, which speeds
up expansion of large data frames significantly. We have added an alert
message for data frames with at least 1000 rows.Minor changes:
unnest2
: Reworked internal code on SplitDFrame
that now only attempts to
recycle rows when length is greater than 1. Additionally, the function now
consistently sets NA
when length is 0. Removed unnecessary internal call
to unlist
-- we can just access the values directly.New functions:
unnest2
: This is a simpler variant of tidyr unnest
that is intended to
work directly on S4 DFrame
without coercion to data.frame
. It supports
unnesting of a single list-column by design, and is intended to be as
simple as possible.Major changes:
cast
: No longer requires tidyr internally. We reworked the internal code of
this function to use base R syntax with full Bioconductor S4 class support.
No internal coercion to data.frame
is performed.Major changes:
Minor changes:
cast
: Now using tidy-select internally with {{
. This is now required for
breaking change update with tidyselect 1.2.0.Minor changes:
melt
: Need to dynamically use Matrix rowSums
generic internally when
applicable, to avoid issues parsing sparseMatrix
objects.Minor changes:
DFrame
instead of DataFrame
virtual class.Minor changes:
rowSums
into NAMESPACE, which has been removed from
BiocGenerics in Bioconductor 3.17 update.Minor changes:
rbindToDataFrame
: Simplified internal construction of DataFrame
from
list
, using new
instead of matrix
coercion internally. This same
approach is now used for as.DataFrame
list
method in pipette.leftJoin
and other join functions: Improved error message handling when
x
and y
contain overlapping columns not defined in by
argument.Major changes:
rbindToDataFrame
: Improved nesting of list columns.Major changes:
mapToDataFrame
, in favor of only supporting rbindToDataFrame
function. This helps lighten the package a bit by removing suggested
purrr and dplyr packages.Minor changes:
rbindToDataFrame
: Simplified internal logic to speed up function quite a
bit. Now supports processing of "cellosaurus.txt"
file.Major changes:
rbindToDataFrame
: Reworked internal code. For very large nested lists,
consider using data.table rbindlist
instead, which is very performant.Minor changes:
NA
values defined in "by"
argument.requireNamespaces
internally (from goalie) for optional package
checks.New functions:
cast
: Added support for casting DataFrame
objects in long format back to
wide format. Internally this calls tidyr::pivot_wider
currently. The
function is intentionally designed to be simple and provide easy
interconversion support for melt
return in long format back to wide format.
For more advanced uses, call pivot_wider
directly instead.Minor changes:
Major changes:
Suggests
instead of Imports
.Major changes:
Minor changes:
DataFrame
function.Minor changes:
unlistToDataFrame
function from NAMESPACE.Minor changes:
abort
function internally instead of stop
for better stylized
error messages.Minor changes:
as.DataFrame
from pipette to improve consistency of list
to DataFrame
coercion, which is currently tricky on Bioconductor 3.13.Minor changes:
mapToDataFrame
: Hardened function in the case when dplyr is not installed.
Internally the purrr::map_df
call hands off to dplyr.Minor changes:
mapToDataFrame
: Offloading purrr as a suggested package, since this is the
only function that calls map_dfr
internally.map_df
and map_dfr
as reexported functions.Minor changes:
leftJoin
: Improve error message when user attempts to pass in columns via
"by" argument that are not unique.Minor changes:
DataFrame
, to avoid unwanted issues with
classed objects.fullJoin
return now retains metadata from x
, as expected.Minor changes:
mutateAll
: Attempting to remove dependency on internal as_tibble
method,
so we can preserve complex S4 columns.Minor changes:
New functions:
rbindToDataFrame
, which is an incredibly useful utility for ensuring
that nested list elements return 1:1 per row. The function handles complex
S4 classes such as IntegerList
and nested ranges.Minor changes:
unlistToDataFrame
to mapToDataFrame
, which better matches the
conventions used in other Acid Genomics packages, as well as purrr.Minor changes:
unlistToDataFrame
: Improved internal code to handle input of CharacterList
and IntegerList
properly. Reworked internal call to purrr::map_dfr
.Minor changes:
Minor changes:
Minor changes:
New functions:
collapseToString
from basejump.Minor changes:
Minor changes:
New functions:
unlistToDataFrame
utility, inspired by approach used in purrr
package. Note that package now imports purrr.Minor changes:
mutateAt
: Bug fix to improve internal list
to DataFrame
coercion.Minor changes:
melt
: Improved internal code for contingency table
method support.Minor changes:
melt
: Added back table
method support, which is used in pointillism.New functions:
melt
, mutateAll
,
mutateIf
, transmuteAt
, transmuteIf
, mutateAt
, selectIf
.Minor changes:
melt
: Restricted method to only support matrix
and DataFrame
here.
Additional S4 methods that work on SummarizedExperiment
and
SingleCellExperiment
are defined in basejump.Initial release.
New functions:
innerJoin
, leftJoin
,
rightJoin
, fullJoin
, semiJoin
, antiJoin
.Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.