@@ -14,12 +14,10 @@ Pre-1.0.0 numbering scheme: 0.x will indicate releases, while 0.x.y will indicat
1414 with ` covid ` . The data set previously named ` jhu_confirmed_cumulative_num ` has
1515 been removed from the package, but a renamed version is has been removed from
1616 the package, but a renamed version is still available in ` epidatasets ` .
17-
18- ## Bug fixes
19-
20- - Removed ` .window_size = 1 ` default from ` epi_slide_{mean,sum,opt} ` ; this
21- argument is now mandatory, and should nearly always be greater than 1 except
22- for testing purposes.
17+ - ` epi_slide_{sum,mean,opt} ` have improved default output column names, and
18+ additional arguments for specifying names: ` .prefix ` , ` .suffix ` ,
19+ ` .new_col_names ` . To obtain the old naming behavior, use `.prefix =
20+ "slide_value_ "`.
2321
2422## Improvements
2523
@@ -29,6 +27,18 @@ Pre-1.0.0 numbering scheme: 0.x will indicate releases, while 0.x.y will indicat
2927- Improved validation of ` .window_size ` arguments.
3028- Rewrote a lot of the package documentation to be more consistent and
3129 informative. Simplified and streamlined the vignettes.
30+ - ` epi_slide_{sum,mean,opt} ` on ungrouped ` epi_df ` s will now temporarily group
31+ by ` geo_value ` and any ` other_keys ` for the slide operation rather than raise
32+ an error about duplicated time values. ` epi_slide ` 's analogous automatic
33+ grouping has been made temporary in order to match.
34+
35+ ## Bug fixes
36+
37+ - Removed ` .window_size = 1 ` default from ` epi_slide_{mean,sum,opt} ` ; this
38+ argument is now mandatory, and should nearly always be greater than 1 except
39+ for testing purposes.
40+ - Fixed ` epi_slide_{sum,mean,opt} ` raising an error on certain tidyselect
41+ expressions.
3242
3343## Cleanup
3444
0 commit comments