fix: Automatically handle datetime conversion for dataset filtering functions

R/datasets.R

@@ -444,6 +444,31 @@ keep_columns_from_dataset <- function(dataset, columns, validate = FALSE) {
   dataset$keep_columns(columns, validate)
 }
 
+#' Convert R datetime objects to Python datetime.datetime objects
+#' 
+#' Wrapped in suppressWarnings() due to unnecessary warning in timedatectl
+#' package: https://stat.ethz.ch/pipermail/r-devel/2018-May/076163.html
+#' 
+#' @param posix_date The POSIX* object to be converted
+#' @return A Python datetime.datetime object
+#' @md
+.posixct_to_datetime <- function(posix_date) {
+  suppressWarnings({
+  datetime <- import("datetime", convert=FALSE)
+
+  parsed_date <- sapply(stringr::str_split(posix_date, "[- : ]")[[1]],
+                        as.integer)
+  names(parsed_date) <- c("year", "month", "day", "hour", "minute", "second")
+
+  datetime$datetime(parsed_date[["year"]],
+                    parsed_date[["month"]],
+                    parsed_date[["day"]],
+                    parsed_date[["hour"]],
+                    parsed_date[["minute"]],
+                    parsed_date[["second"]])
+  })
+}
+
 #' Filter Tabular Dataset with time stamp columns after a specified start time.
 #'
 #' @description
@@ -458,6 +483,7 @@ keep_columns_from_dataset <- function(dataset, columns, validate = FALSE) {
 #' @md
 filter_dataset_after_time <- function(dataset, start_time,
                                       include_boundary = TRUE) {
+  start_time <- .posixct_to_datetime(start_time)
   dataset$time_after(start_time, include_boundary)
 }
 
@@ -475,6 +501,7 @@ filter_dataset_after_time <- function(dataset, start_time,
 #' @md
 filter_dataset_before_time <- function(dataset, end_time,
                                        include_boundary = TRUE) {
+  end_time <- .posixct_to_datetime(end_time)
   dataset$time_before(end_time, include_boundary)
 }
 
@@ -493,6 +520,8 @@ filter_dataset_before_time <- function(dataset, end_time,
 #' @md
 filter_dataset_between_time <- function(dataset, start_time, end_time,
                                         include_boundary = TRUE) {
+  start_time <- .posixct_to_datetime(start_time)
+  end_time <- .posixct_to_datetime(end_time)
   dataset$time_between(start_time, end_time, include_boundary)
 }
 

tests/testthat/test_datasets.R

@@ -10,11 +10,14 @@ test_that("create a tabular dataset,
   ws <- existing_ws
 
   # create tabular dataset from delimited files
+  date <- as.POSIXct("2011-05-01 17:55:23")
   path_to_dataset <- "https://automlsamplenotebookdata.blob.core.windows.net/automl-sample-notebook-data/nyc_energy.csv"
-  dataset <- create_tabular_dataset_from_delimited_files(path=path_to_dataset)
-
+  time_column_name <- 'timeStamp'
+  dataset <- create_tabular_dataset_from_delimited_files(path=path_to_dataset)$with_timestamp_columns(fine_grain_timestamp=time_column_name)
+  filtered_dataset <- filter_dataset_before_time(dataset, date)
+
   # load data into data frame
-  pandas_df <- load_dataset_into_data_frame(dataset)
+  pandas_df <- load_dataset_into_data_frame(filtered_dataset)
   expect_equal(is.data.frame(pandas_df), TRUE)
 
   # register first version of the dataset
@@ -28,7 +31,7 @@ test_that("create a tabular dataset,
   expect_equal(registered_dataset1$name, registered_dataset2$name)
   expect_equal(registered_dataset1$description, 'I am version 1')
   expect_equal(registered_dataset2$description, 'I am version 2')
-                   
+
   # unregister datasets
   unregister_all_dataset_versions(dataset)
   expect_equal(dataset$name, NULL)