[GitHub] [arrow] thisisnic commented on a change in pull request #10507: ARROW-13022: [R] bindings for lubridate's year, isoyear, quarter, month, day, wday, yday, isoweek, minute, and second functions

[GitBox <gi...@apache.org>]

thisisnic commented on a change in pull request #10507:
URL: https://github.com/apache/arrow/pull/10507#discussion_r649907028



##########
File path: r/R/dplyr-functions.R
##########
@@ -442,3 +442,37 @@ nse_funcs$strptime <- function(x, format = "%Y-%m-%d %H:%M:%S", tz = NULL, unit
 
   Expression$create("strptime", x, options = list(format = format, unit = unit))
 }
+
+nse_funcs$wday <- function(x, label = FALSE, abbr = TRUE, week_start = getOption("lubridate.week.start", 7)) {
+  if (label) {
+    arrow_not_supported("Label argument")
+  }
+  offset <- get_date_offset(week_start)
+  Expression$create("add", Expression$create("day_of_week", x), Expression$scalar(offset))
+}
+
+#' Get date offset
+#' 
+#' Arrow's `day_of_week` kernel counts from 0 (Monday) to 6 (Sunday), whereas
+#' `lubridate::wday` counts from 1 to 7, and allows users to specify which day
+#' of the week is first (Sunday by default).  This function converts the returned

Review comment:
       That probably makes more sense than my workaround here actually.

##########
File path: r/R/expression.R
##########
@@ -28,8 +28,17 @@
   # stringr spellings of those
   "str_length" = "utf8_length",
   "str_to_lower" = "utf8_lower",
-  "str_to_upper" = "utf8_upper"
+  "str_to_upper" = "utf8_upper",
   # str_trim is defined in dplyr.R
+  "year" = "year",
+  "isoyear" = "iso_year",
+  "quarter" = "quarter",
+  "month" = "month",
+  "day" = "day",
+  "yday" = "day_of_year",
+  "isoweek" = "iso_week",
+  "minute" = "minute",
+  "second" = "second"

Review comment:
       Good catch.  Hmm, I think this might be more complicated than that actually, as lubridate rounds it to 1 decimal place, using R's odd-even rounding whereas Arrow doesn't do the rounding, so I'm not sure if this can be done without rounding being implemented (see https://issues.apache.org/jira/browse/ARROW-12744)

##########
File path: r/R/expression.R
##########
@@ -28,8 +28,17 @@
   # stringr spellings of those
   "str_length" = "utf8_length",
   "str_to_lower" = "utf8_lower",
-  "str_to_upper" = "utf8_upper"
+  "str_to_upper" = "utf8_upper",
   # str_trim is defined in dplyr.R
+  "year" = "year",
+  "isoyear" = "iso_year",
+  "quarter" = "quarter",
+  "month" = "month",
+  "day" = "day",
+  "yday" = "day_of_year",
+  "isoweek" = "iso_week",
+  "minute" = "minute",

Review comment:
       Oops, will add in now.

##########
File path: r/R/dplyr-functions.R
##########
@@ -442,3 +442,37 @@ nse_funcs$strptime <- function(x, format = "%Y-%m-%d %H:%M:%S", tz = NULL, unit
 
   Expression$create("strptime", x, options = list(format = format, unit = unit))
 }
+
+nse_funcs$wday <- function(x, label = FALSE, abbr = TRUE, week_start = getOption("lubridate.week.start", 7)) {
+  if (label) {
+    arrow_not_supported("Label argument")
+  }
+  offset <- get_date_offset(week_start)
+  Expression$create("add", Expression$create("day_of_week", x), Expression$scalar(offset))
+}
+
+#' Get date offset
+#' 
+#' Arrow's `day_of_week` kernel counts from 0 (Monday) to 6 (Sunday), whereas
+#' `lubridate::wday` counts from 1 to 7, and allows users to specify which day
+#' of the week is first (Sunday by default).  This function converts the returned

Review comment:
       OK, will do, and will open a JIRA, thanks!

##########
File path: r/R/dplyr-functions.R
##########
@@ -442,3 +442,37 @@ nse_funcs$strptime <- function(x, format = "%Y-%m-%d %H:%M:%S", tz = NULL, unit
 
   Expression$create("strptime", x, options = list(format = format, unit = unit))
 }
+
+nse_funcs$wday <- function(x, label = FALSE, abbr = TRUE, week_start = getOption("lubridate.week.start", 7)) {
+  if (label) {
+    arrow_not_supported("Label argument")
+  }
+  offset <- get_date_offset(week_start)
+  Expression$create("add", Expression$create("day_of_week", x), Expression$scalar(offset))
+}
+
+#' Get date offset
+#' 
+#' Arrow's `day_of_week` kernel counts from 0 (Monday) to 6 (Sunday), whereas
+#' `lubridate::wday` counts from 1 to 7, and allows users to specify which day
+#' of the week is first (Sunday by default).  This function converts the returned

Review comment:
       https://issues.apache.org/jira/browse/ARROW-13054

##########
File path: r/R/expression.R
##########
@@ -28,8 +28,17 @@
   # stringr spellings of those
   "str_length" = "utf8_length",
   "str_to_lower" = "utf8_lower",
-  "str_to_upper" = "utf8_upper"
+  "str_to_upper" = "utf8_upper",
   # str_trim is defined in dplyr.R
+  "year" = "year",
+  "isoyear" = "iso_year",
+  "quarter" = "quarter",
+  "month" = "month",
+  "day" = "day",
+  "yday" = "day_of_year",
+  "isoweek" = "iso_week",
+  "minute" = "minute",
+  "second" = "second"

Review comment:
       So it does; I think I must have been getting mixed up with something I'd changed when I needed to update a test I wrote, never mind,

##########
File path: r/tests/testthat/test-dplyr-lubridate.R
##########
@@ -0,0 +1,119 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+library(lubridate)
+library(dplyr)
+
+test_date <- ymd_hms("1987-10-09 23:00:00", tz = NULL)
+test_df <- tibble::tibble(date = test_date)
+
+test_that("extract datetime components from date", {
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = year(date)) %>%
+      collect(),
+    test_df
+  )
+  
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = isoyear(date)) %>%
+      collect(),
+    test_df
+  )
+  
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = quarter(date)) %>%
+      collect(),
+    test_df
+  )
+  
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = month(date)) %>%
+      collect(),
+    test_df
+  )
+  
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = wday(date)) %>%
+      collect(),
+    test_df
+  )
+  
+  expect_dplyr_equal(
+    input %>%
+      mutate(x = wday(date, week_start = 3)) %>%
+      collect(),
+    test_df
+  )
+  
+  expect_warning(
+    test_df %>%
+      Table$create() %>%
+      mutate(x = wday(date, label = TRUE)) %>%
+      collect(),
+    regexp = "Label argument not supported by Arrow; pulling data into R"
+  )
+  
+  expect_warning(
+    test_df %>%
+      Table$create() %>%
+      mutate(x = wday(date, locale = Sys.getlocale("LC_TIME"))) %>%
+      collect(),
+    regexp = 'Expression wday(date, locale = Sys.getlocale("LC_TIME")) not supported in Arrow; pulling data into R',
+    fixed = TRUE
+  )

Review comment:
       Good point, I think I was just copying and pasting from the older way of doing things.  Whether it matters or not to @nealrichardson, I think that way is a lot clearer so I'll update them!

##########
File path: r/R/dplyr-functions.R
##########
@@ -442,3 +442,37 @@ nse_funcs$strptime <- function(x, format = "%Y-%m-%d %H:%M:%S", tz = NULL, unit
 
   Expression$create("strptime", x, options = list(format = format, unit = unit))
 }
+
+nse_funcs$wday <- function(x, label = FALSE, abbr = TRUE, week_start = getOption("lubridate.week.start", 7)) {
+  if (label) {
+    arrow_not_supported("Label argument")

Review comment:
       The label argument is whether the day is displayed as a character string or a number.  Maybe it is something we should support; shall I open a ticket?




-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org