move news

NEWS.md

@@ -26,6 +26,9 @@
 
 7. `fread`'s `fill` argument now also accepts an `integer` in addition to boolean values. `fread` always guesses the number of columns based on reading a sample of rows in the file. When `fill=TRUE`, `fread` stops reading and ignores subsequent rows when this estimate winds up too low, e.g. when the sampled rows happen to exclude some rows that are even wider, [#2727](https://github.com/Rdatatable/data.table/issues/2727) [#2691](https://github.com/Rdatatable/data.table/issues/2691) [#4130](https://github.com/Rdatatable/data.table/issues/4130) [#3436](https://github.com/Rdatatable/data.table/issues/3436). Providing an `integer` as argument for `fill` allows for a manual estimate of the number of columns instead, [#1812](https://github.com/Rdatatable/data.table/issues/1812) [#5378](https://github.com/Rdatatable/data.table/issues/5378). Thanks to @jangorecki, @christellacaze, @Yiguan, @alexdthomas, @ibombonato, @Befrancesco, @TobiasGold for reporting/requesting, and Benjamin Schwendinger for the PR.
 
+8. `rbindlist(l, use.names=TRUE)` and `rbind` allows now the automatic binding of columns with different class attributes for certain classes such as `Date`, `IDate`, `ITime`, `POSIXct` and `AsIs` with other columns of similar classes, e.g., `IDate` and `Date`. The conversion is done automatically and the class attribute of the final column is determined by the first encountered class attribute in the binding list, [#5309](https://github.com/Rdatatable/data.table/issues/5309), [#4934](https://github.com/Rdatatable/data.table/issues/4934), [#5391](https://github.com/Rdatatable/data.table/issues/5391).
+`rbindlist(l, ignore.attr=TRUE)` and `rbind` also gained argument `ignore.attr` to manually deactivate the safety-net of binding columns with different column classes, [#3911](https://github.com/Rdatatable/data.table/issues/3911), [#5542](https://github.com/Rdatatable/data.table/issues/5542). Thanks to @dcaseykc, @fox34, @adrian-quintario, @berg-michael, @arunsrinivasan, @statquant, @pkress, @jrausch12, @therosko, @OfekShilon, @iMissile, @tdhock for the request and @ben-schwen for the PR.
+
 ## BUG FIXES
 
 1. `unique()` returns a copy the case when `nrows(x) <= 1` instead of a mutable alias, [#5932](https://github.com/Rdatatable/data.table/pull/5932). This is consistent with existing `unique()` behavior when the input has no duplicates but more than one row. Thanks to @brookslogan for the report and @dshemetov for the fix.
@@ -347,8 +350,6 @@
 
 41. `tables()` is faster by default by excluding the size of character strings in R's global cache (which may be shared) and excluding the size of list column items (which also may be shared). `mb=` now accepts any function which accepts a `data.table` and returns a higher and better estimate of its size in bytes, albeit more slowly; e.g. `mb = utils::object.size`.
 
-43. `rbindlist(l, use.names=TRUE)` and `rbind` allows now the automatic binding of columns with different class attributes for certain classes such as `Date`, `IDate`, `ITime`, `POSIXct` and `AsIs` with other columns of similar classes, e.g., `IDate` and `Date`. The conversion is done automatically and the class attribute of the final column is determined by the first encountered class attribute in the binding list, [#5309](https://github.com/Rdatatable/data.table/issues/5309), [#4934](https://github.com/Rdatatable/data.table/issues/4934), [#5391](https://github.com/Rdatatable/data.table/issues/5391).
-`rbindlist(l, ignore.attr=TRUE)` and `rbind` also gained argument `ignore.attr` to manually deactivate the safety-net of binding columns with different column classes, [#3911](https://github.com/Rdatatable/data.table/issues/3911), [#5542](https://github.com/Rdatatable/data.table/issues/5542). Thanks to @dcaseykc, @fox34, @adrian-quintario, @berg-michael, @arunsrinivasan, @statquant, @pkress, @jrausch12, @therosko, @OfekShilon, @iMissile, @tdhock for the request and @ben-schwen for the PR.
 
 ## BUG FIXES