1cadfc3 | Romain Francois | 28 January 2020, 07:39:42 UTC | update error after previous reveiw commit | 28 January 2020, 07:39:42 UTC |
7af73e8 | Romain François | 28 January 2020, 07:29:46 UTC | Update R/conditions.R Co-Authored-By: Hadley Wickham <h.wickham@gmail.com> | 28 January 2020, 07:29:46 UTC |
b0358c4 | Romain Francois | 27 January 2020, 18:18:30 UTC | tweaks to arrange() errors | 27 January 2020, 18:18:30 UTC |
46820b6 | Hadley Wickham | 25 January 2020, 14:18:25 UTC | Unpack tibbles in bind_cols() Fixes #4804 | 25 January 2020, 14:18:25 UTC |
dbb5acc | Hadley Wickham | 24 January 2020, 23:23:43 UTC | Restore list-binding behaviour Fixes #4697 | 24 January 2020, 23:27:47 UTC |
7a832ba | Hadley Wickham | 24 January 2020, 22:00:40 UTC | Correct binds test file name | 24 January 2020, 23:27:24 UTC |
42e203a | Hadley Wickham | 24 January 2020, 22:42:28 UTC | Carefully simplify bind implementations Preserving existing behaviour as much as possible. | 24 January 2020, 23:27:01 UTC |
6d2d546 | Hadley Wickham | 23 January 2020, 21:36:16 UTC | Weaken list_of() tests so they pass with dev vctrs | 23 January 2020, 21:36:16 UTC |
bdb40b8 | Hadley Wickham | 23 January 2020, 18:21:12 UTC | Re-document | 23 January 2020, 18:21:12 UTC |
273d8f8 | Hadley Wickham | 23 January 2020, 15:17:27 UTC | Polishing news | 23 January 2020, 15:17:27 UTC |
13e4bae | Hadley Wickham | 23 January 2020, 15:07:47 UTC | Divide news into groups | 23 January 2020, 15:07:47 UTC |
baf1168 | Hadley Wickham | 23 January 2020, 14:55:42 UTC | Control position of new mutate() columns (#4774) Fixes #2047 | 23 January 2020, 14:55:42 UTC |
0338796 | Hadley Wickham | 22 January 2020, 20:58:48 UTC | Test 0-row group case | 22 January 2020, 20:58:48 UTC |
6d916ed | Hadley Wickham | 22 January 2020, 20:55:52 UTC | Continue to refactor summarise() tests | 22 January 2020, 20:55:52 UTC |
4805d46 | Hadley Wickham | 22 January 2020, 20:37:14 UTC | More re-org and tweaking | 22 January 2020, 20:37:14 UTC |
34baa87 | Hadley Wickham | 22 January 2020, 20:16:39 UTC | Roughly group summarise tests | 22 January 2020, 20:16:39 UTC |
6bb8d89 | Hadley Wickham | 22 January 2020, 20:09:24 UTC | Remove summarise tests I've removed tests that seem specifically related to hybrid eval or vctrs | 22 January 2020, 20:12:28 UTC |
7470ed0 | Hadley Wickham | 22 January 2020, 15:58:02 UTC | Give more control over which variables mutate() should keep (#4773) Adds new .keep argument with options "none", "used", "unused", and "all". Fixes #3721 | 22 January 2020, 15:58:02 UTC |
08784ae | Hadley Wickham | 22 January 2020, 15:24:48 UTC | Redocument | 22 January 2020, 15:24:48 UTC |
abe8b84 | Hadley Wickham | 22 January 2020, 15:22:22 UTC | Skip failing test for now | 22 January 2020, 15:22:22 UTC |
1b82854 | Hadley Wickham | 22 January 2020, 15:19:48 UTC | Simplify glue_c() | 22 January 2020, 15:19:48 UTC |
6f05ed2 | Romain François | 21 January 2020, 21:07:14 UTC | Add names= argument to across() (#4782) * + across(names=) closes #4781 * use names= as argument of across(), not summarise() * missing `)` in examples * initial doc for across(names=) * Update R/across.R Co-Authored-By: Hadley Wickham <h.wickham@gmail.com> * Update R/across.R Co-Authored-By: Hadley Wickham <h.wickham@gmail.com> * using expect_named() for across tests. simplify implementation of across() so that all goes through the same code. * update doc for return of across() * using %||% * Simplify across() impl and tests. Co-authored-by: Hadley Wickham <h.wickham@gmail.com> | 21 January 2020, 21:07:14 UTC |
9358fc5 | Romain François | 21 January 2020, 14:26:49 UTC | test arrange() support for across() (#4786) closes #4679 | 21 January 2020, 14:26:49 UTC |
e0daacc | Romain François | 21 January 2020, 14:23:49 UTC | row wise specific error bullets (#4783) * Fix stop_filter_incompatible_size() display. * Add hint to enlist in mutate incompatible size exception * Hint to enlist with summarise(<rowwise>) on unsupported type error * refactoring into cnd_bullet_rowwise_unlist() * Using friendly_type_of() instead of direct as_friezndly_type() * errors changed in newest tidyselect | 21 January 2020, 14:23:49 UTC |
aea647e | Hiroaki Yutani | 21 January 2020, 08:51:32 UTC | Fix a typo (#4785) | 21 January 2020, 08:51:32 UTC |
ad0e038 | Romain François | 20 January 2020, 16:56:57 UTC | Detailed information about the keys of a group on error (#4746) * Rework stop_eval_tidy() to be able to also use it with summarise() and other functions * custom errors for incompatible sizes * Also promote vctrs_error so that they have more context * custom handling of the error produced by vec_c(!!!chunks) * ..{index} is, and headeing for mixed types * no backticks around typeof() result * rm the redundant `= NULL` test * use name instead of ..{index} when possible * Another ..{index} is * An earlier column had size * Applying comments from @hadley review * move the tryCatch() around the for() loop * Added "or 1" since summarise() now supports tidy recycling * Detailed information about the keys of a group on error * Need repeating "i" * update error output * update bullets * squish group info into a single bullet * stress error group info with dates and lists * Give up some tests because of this problem https://github.com/r-lib/testthat/issues/978 * Apply group_info() to other errors * using cur_group_id() and cur_group() * Simplify group_info() * inline tibble:::format_v * + cur_group_label() internal generic to generate the label for the current group * Not reporting group information on combine errors from vec_c() * Correctly report faulty group in summarise() recycling problem * Detail difference of vctr types for each group * set_error_context(), cnd_problem() and cnd_bullet_current_expression() * Refactor into stop_dplyr() * quote expression * report correct faulty group in mutate() on recycling problems. * mention result name * Simplify glue_c() * move cur_group_label() to context.R * Simplify stop_mutate_recycle_incompatible_size() * Species = "setosa" * + tests showing fix for "1 or 1" situation | 20 January 2020, 16:56:57 UTC |
685eebd | Hadley Wickham | 20 January 2020, 13:16:13 UTC | Rename dplyr_tidy_eval to dplyr_data_masking | 20 January 2020, 13:16:24 UTC |
df25ee9 | Ian Cook | 20 January 2020, 13:06:35 UTC | Implement keep = T in left_join(), right_join() (#4762) | 20 January 2020, 13:06:35 UTC |
1614550 | Hadley Wickham | 20 January 2020, 12:52:40 UTC | Remove out dated article | 20 January 2020, 12:52:40 UTC |
4033be3 | Hadley Wickham | 19 January 2020, 15:43:47 UTC | Simplify filter C code By using context functions in R | 19 January 2020, 15:47:12 UTC |
7d02465 | Hadley Wickham | 19 January 2020, 15:37:40 UTC | Simplify filter() C implementation Since regrouping is now handled similarly across all row slicing verbs | 19 January 2020, 15:37:40 UTC |
d1c88bc | Hadley Wickham | 19 January 2020, 15:25:37 UTC | Improvements to contextual functions (#4772) * New `cur_data()` and `cur_group_rows() * Connect contextual functions to data.table pronouns * Access `cur_group_id()` and `n()` through mask instead of context Fixes #4764 | 19 January 2020, 15:25:37 UTC |
afb7130 | Hadley Wickham | 18 January 2020, 16:23:21 UTC | New programming with dplyr vignette (#4724) Fixes #4407 | 18 January 2020, 16:23:21 UTC |
c15eefa | Hadley Wickham | 18 January 2020, 16:09:06 UTC | Work around buglet in rlang | 18 January 2020, 16:09:31 UTC |
9508c7b | Lionel Henry | 18 January 2020, 11:50:31 UTC | Fix defusal of tidyselect arguments cc @hadley | 18 January 2020, 11:50:31 UTC |
33d9fb4 | Hadley Wickham | 17 January 2020, 21:26:28 UTC | Implement relocate Fixes #4598 | 17 January 2020, 21:26:28 UTC |
f37197c | Hadley Wickham | 17 January 2020, 21:23:59 UTC | Redocument | 17 January 2020, 21:23:59 UTC |
205386c | Hadley Wickham | 17 January 2020, 20:17:33 UTC | Move DataMask to data-mask.R | 17 January 2020, 20:17:33 UTC |
bb38491 | Hadley Wickham | 17 January 2020, 15:52:45 UTC | Mark scoped helpers as superseded (#4769) | 17 January 2020, 15:52:45 UTC |
dd260ba | Hadley Wickham | 17 January 2020, 15:32:25 UTC | Improve mutate()/summarise() multi-row/multi-col docs Fixes #4721 | 17 January 2020, 15:32:25 UTC |
0170d18 | Hadley Wickham | 17 January 2020, 14:36:18 UTC | Emphasise that rowwise(...) is different to group_by(...) Fixes #4768 | 17 January 2020, 14:36:18 UTC |
3e8e17d | Hadley Wickham | 17 January 2020, 14:35:04 UTC | mutate() doc tweaks | 17 January 2020, 14:35:04 UTC |
5f4c45a | Hadley Wickham | 17 January 2020, 14:26:11 UTC | Tweak reference index for condense/rowwise | 17 January 2020, 14:26:11 UTC |
4cc84d4 | Hadley Wickham | 17 January 2020, 14:13:59 UTC | Use lowercase rd for pkgdown compatibility | 17 January 2020, 14:13:59 UTC |
2b25da1 | Hadley Wickham | 17 January 2020, 13:28:38 UTC | with_groups is experimental | 17 January 2020, 13:28:38 UTC |
bb1c69c | Hadley Wickham | 17 January 2020, 13:24:24 UTC | Explain experimental/questioning/superseded | 17 January 2020, 13:24:24 UTC |
bc80294 | Hadley Wickham | 17 January 2020, 13:08:53 UTC | WS | 17 January 2020, 13:08:53 UTC |
1aca4c7 | Hadley Wickham | 17 January 2020, 13:04:50 UTC | Consistently place badge at top of description And remove some extraneous whitespace | 17 January 2020, 13:04:50 UTC |
ea7f5e9 | Romain François | 17 January 2020, 12:54:30 UTC | Revised rowwise() + condense() (#4760) * rowwise() comes back to life, optionally preserves variables when summarising, and maintains grouping by row through all verb transformations * condense() is a new form of summarise() that always returns a single row, by wrapping each new variable in a list. * do() is deprecated Fixes #4723 | 17 January 2020, 12:54:30 UTC |
057b453 | Romain François | 16 January 2020, 16:49:20 UTC | Fix cur_group() (#4758) * Fix DataMask$current_key() :grimacing: * + test for cur_group() closes #4750 * test the results of cur_group() directly, i.e. no combining and auto splicing post summarise() | 16 January 2020, 16:49:20 UTC |
2cfdfd9 | Hadley Wickham | 16 January 2020, 11:39:27 UTC | News tweaking | 16 January 2020, 11:46:21 UTC |
b7c5e04 | Hadley Wickham | 15 January 2020, 23:50:00 UTC | Document slice helpers with slice | 16 January 2020, 11:46:21 UTC |
ccb5256 | Hadley Wickham | 15 January 2020, 23:49:19 UTC | Improve section names | 16 January 2020, 11:46:21 UTC |
80ae428 | Hadley Wickham | 15 January 2020, 20:06:46 UTC | Simplify pkgconfig usage Add deliberately reduce advertisement of it, since we don't use it elsewhere, and aren't planning to. | 16 January 2020, 11:46:21 UTC |
bdfd09f | Hadley Wickham | 15 January 2020, 16:55:08 UTC | Document join methods with generics | 16 January 2020, 11:46:21 UTC |
3fa41f6 | Romain François | 15 January 2020, 20:37:53 UTC | reduce redundant code with a few macros (#4756) * reduce redundant code with a few macros * internal mutate() and summarise() code does not need `dots_name` or `i` | 15 January 2020, 20:37:53 UTC |
c58f91b | Romain François | 15 January 2020, 16:45:13 UTC | bullet errors for mutate() (#4747) Improve error messages for mutate() | 15 January 2020, 16:45:13 UTC |
6daea45 | Hadley Wickham | 15 January 2020, 16:15:49 UTC | Better documentation for generic functions (#4689) Using a new helper that automatically links to methods in loaded packages. Note that it does not work when previewing dev documentation in RStudio, because RStudio currently renders in a separate R process Fixes #4688 | 15 January 2020, 16:15:49 UTC |
2de2523 | Hadley Wickham | 15 January 2020, 16:15:36 UTC | Merge remote-tracking branch 'refs/remotes/origin/master' Conflicts: man/arrange.Rd | 15 January 2020, 16:15:36 UTC |
f1bf515 | Hadley Wickham | 15 January 2020, 16:15:00 UTC | Polish titles and of primary verbs | 15 January 2020, 16:15:00 UTC |
ed8647b | Hadley Wickham | 15 January 2020, 15:29:24 UTC | Apply pattern to all major verbs | 15 January 2020, 15:29:24 UTC |
2fb38e5 | Hadley Wickham | 15 January 2020, 15:15:07 UTC | Don't link to self | 15 January 2020, 15:15:07 UTC |
3e36daa | Hadley Wickham | 15 January 2020, 15:11:17 UTC | Add extra line break | 15 January 2020, 15:11:17 UTC |
b693e92 | Hadley Wickham | 15 January 2020, 15:09:16 UTC | Note about RStudio's behaviour | 15 January 2020, 15:09:16 UTC |
3ccfb79 | Hadley Wickham | 15 January 2020, 15:08:23 UTC | Inherit params from arrange() Since it's alphabetically first | 15 January 2020, 15:08:23 UTC |
6265622 | Hadley Wickham | 15 January 2020, 15:05:07 UTC | Merge remote-tracking branch 'refs/remotes/origin/master' Conflicts: R/manip.r | 15 January 2020, 15:05:07 UTC |
7756858 | Hadley Wickham | 15 January 2020, 14:51:19 UTC | Revise old tidy eval docs * Remove old explanations * Point to latest resources * Link on to rlang docs * Don't need individual imports since we import the whole package * Drop .data alias, since we can (because it starts with `.`?) Part of #3734 | 15 January 2020, 14:51:47 UTC |
23ad9cc | Hadley Wickham | 15 January 2020, 14:00:51 UTC | Fix thinko in generics doc | 15 January 2020, 14:00:51 UTC |
0f06b30 | Hadley Wickham | 15 January 2020, 14:00:43 UTC | Tease apart documentation for join functions | 15 January 2020, 14:00:43 UTC |
c352300 | Hadley Wickham | 15 January 2020, 13:25:24 UTC | Eliminate expect_groups() helper | 15 January 2020, 13:25:24 UTC |
9bca9ba | Hadley Wickham | 15 January 2020, 13:18:23 UTC | Mark all_equal() as questioning And rename test to match R/ file name | 15 January 2020, 13:18:23 UTC |
edd488f | Hadley Wickham | 15 January 2020, 13:12:04 UTC | Separate select() and rename() functions Including documentation and tests | 15 January 2020, 13:12:04 UTC |
6c912be | Hadley Wickham | 15 January 2020, 12:55:48 UTC | Refine package docs | 15 January 2020, 12:55:48 UTC |
56c2197 | Hadley Wickham | 15 January 2020, 12:46:14 UTC | Minimise regrouping work (#4751) This PR starts to develop the dplyr "interface", i.e. the set of generics that you need to provide methods for if you want to extend dplyr to work with new data frame subclasses. It also uses those methods (along with a count_regroups()) to ensure that the existing grouped_df implementations are not needlessly regrouping data. Fixes #4086 because count() can now use dplyr_reconstruct() to restore the original class Fixes #4051 because I've carefully documented the return value of the major verbs Fixes #4711 since implementing with_groups() is now easy. | 15 January 2020, 12:46:14 UTC |
8df2d66 | Hadley Wickham | 13 January 2020, 23:22:33 UTC | Improve documentation of output type Fixes #4699 | 13 January 2020, 23:22:33 UTC |
3e24c9d | Hadley Wickham | 13 January 2020, 21:53:51 UTC | Remove code doubled via accidental merge | 13 January 2020, 21:53:51 UTC |
3abcf7e | Hadley Wickham | 13 January 2020, 21:51:03 UTC | Consistent names for verb helpers | 13 January 2020, 21:51:03 UTC |
1e76e8a | Hadley Wickham | 13 January 2020, 21:45:39 UTC | Rewrite all join functions (#4741) Now based on a common approach to modifying x rather than creating a new data frame. New `join_rows()` and `join_cols()` provide common toolset for generate row and col indices. New `filter_join()` and `mutate_join()` reduce duplication in joining code. Tests have been completely rewritten from scratch to focus on the concepts of preserving x (type, row order, and column order) that have become more clear. The test for both empty suffixes has been removed. This mildly weakens the guarantees offered by the join functions, but makes `nest_join()` work as it used to. I don't think adding extra logic to make it error in regular joins is worth it. * Fixes #4206: refactoring revealed a trivial implementation (mostly just removing an exiting error message) * Fixes #4225: the data frame method now does all the work * Fixes #4589: refactoring revealed that `nest_join(keep = T)` could share code with `full_join(keep = T)` | 13 January 2020, 21:45:39 UTC |
57f2c27 | Hadley Wickham | 13 January 2020, 21:11:15 UTC | Default cols = everything() So you can use across() to easily access all current data. | 13 January 2020, 21:11:15 UTC |
406a9b7 | Hadley Wickham | 13 January 2020, 15:38:53 UTC | Rename context functions to common prefix (#4748) And add some initial tests | 13 January 2020, 15:38:53 UTC |
90b3e4e | Romain François | 13 January 2020, 14:46:25 UTC | Bullet errors for summarise() (#4729) Better error messages for summarise() | 13 January 2020, 14:46:25 UTC |
887d40b | Kirill Müller | 13 January 2020, 14:38:03 UTC | Set tibble version (#4745) | 13 January 2020, 14:38:03 UTC |
125d75d | Hadley Wickham | 13 January 2020, 14:11:14 UTC | Simplify group metadata (#4728) All metadata functions are now defined and documented together. They provide data.frame methods based on `group_data()`; this will slightly worsen performance but considerably simplifies the implementation. * Use of `...` in `group_keys()` and `group_indices()` is now deprecated to make them pure metadata function * 0-arg form of `group_indices()` (used inside `summarise()`/`mutate()`) is now `cur_group_id()` | 13 January 2020, 14:11:14 UTC |
f7f6682 | Hadley Wickham | 13 January 2020, 14:02:00 UTC | Tweak error message | 13 January 2020, 14:02:00 UTC |
e3ff1f8 | Hadley Wickham | 13 January 2020, 14:00:54 UTC | Simplifying build_groups a little (#4727) * Require vars be a character vector * Move grouped_df() tests out of test-group_by * Move user facing error into group_by_prepare() * Simplify warning code * Simplify group_vars() logic a little * Use errors that work the same across R versions | 13 January 2020, 14:00:54 UTC |
fb39273 | Romain François | 13 January 2020, 06:54:27 UTC | summarise() tidy recycling (#4733) * summarise() obeys tidy recycleing | 13 January 2020, 06:54:27 UTC |
dc8ebd4 | Kirill Müller | 13 January 2020, 00:55:58 UTC | Fix broken test (#4743) | 13 January 2020, 00:55:58 UTC |
5c42cb8 | Hadley Wickham | 12 January 2020, 21:57:48 UTC | Pin tibble to known good version | 12 January 2020, 21:58:04 UTC |
eceeb0f | Romain François | 12 January 2020, 07:40:52 UTC | mutate() can "remove" a variable that did not exist (#4732) * mutate() can remove a variable that did not exist before by setting to NULL. This was tested for but failed. | 12 January 2020, 07:40:52 UTC |
b44c814 | Hadley Wickham | 11 January 2020, 15:09:26 UTC | group_by(.dots) is deprecated #4734 | 11 January 2020, 15:09:26 UTC |
8d7ff18 | Hadley Wickham | 11 January 2020, 15:01:28 UTC | Fix accidental regression Fixes #4734 | 11 January 2020, 15:01:28 UTC |
cab26df | Hadley Wickham | 10 January 2020, 13:49:40 UTC | Remove unnused match_vars() function | 10 January 2020, 13:49:40 UTC |
a2a1031 | Romain François | 10 January 2020, 08:17:37 UTC | stop_dplyr() experiment (#4726) * use `glue_c()` for error messages. | 10 January 2020, 08:17:37 UTC |
168db5d | Hadley Wickham | 09 January 2020, 22:50:40 UTC | Re-document | 09 January 2020, 22:50:40 UTC |
cbbe04e | Hadley Wickham | 09 January 2020, 22:42:06 UTC | Clarify grouped_df() * Make it more clear what the return type is * Rename make_grouped_df_groups_attribute() to compute_groups() | 09 January 2020, 22:42:06 UTC |
df02e7e | Hadley Wickham | 09 January 2020, 22:34:48 UTC | Re-arrange grouped-df.R contents | 09 January 2020, 22:34:48 UTC |
b546893 | Hadley Wickham | 09 January 2020, 22:31:03 UTC | A little tidying of mutate tests | 09 January 2020, 22:31:03 UTC |
258a845 | Hadley Wickham | 09 January 2020, 21:53:28 UTC | Add return type tests | 09 January 2020, 21:53:28 UTC |
98eb4c2 | Hadley Wickham | 09 January 2020, 21:50:53 UTC | Trim and reorganise arrange() tests | 09 January 2020, 21:50:53 UTC |
177db19 | Hadley Wickham | 09 January 2020, 21:26:58 UTC | Verb genericity (#4715) * Genericity for row selection verbs, `filter()`, `arrange()`, `slice()`, `semi_join()`, `anti_join()` This revealed a couple of subtle issues in the tests: the return type of semi_join() and anti_join() can always just come from x, and the error messages needed a conversion to verify_output() * Genericity for transmute and mutate. Fixes #4717 in passing * Start on genericity for summarise() | 09 January 2020, 21:26:58 UTC |