NEWS.md file to track changes to the package.Fixed a bug in the index for empty rows returned by print_report() after remove_constants(). It was always returning the number of rows in the original data frame, instead of the index of empty rows (#173, @Bisaloo).
Fixed a bug in the date guesser used by the standardize_dates() function. The followings have been added or corrected (#175, @Karim-Mane):
Rendering the report no more requires the use of {withr} package. Files generated during this process will be stored in the R temporary directory (#165, @Bisaloo).
Simplification and improvement of code performance across several functions (#154, #156, #162, @Bisaloo).
Replace base R pipe with {tidyverse} pipe (#155, @Karim-Mane)
Stop automatic conversion of numeric values into Date due to unexpected results from date_guess() function (#151, @Karim-Mane).
Allow for a vector of character for na_strings argument in replace_missing_values() (#180, Karim-Mane)
Differentiate message sent when duplicates are found in subject ids column vs duplicates across several columns (#180, Karim-Mane)
Allow for iterative constant data removal (#180, Karim-Mane)
remove_constants() now works fine when the cut-off is different from 1. More importantly, the function no longer depends on {janitor} (#177, Karim-Mane).
scan_data() now makes use of date_guess() to parse date from characters and lubridate::as_date() to parse dates from numeric. This function now operates only on character columns (#181, Karim-Mane).
Replace snakecase::to_snake_case() with janitor::clean_names() and make sure variable names specified in keep and rename are preserved after applying janitor::clean_names() (#180, Karim-Mane).
Function documentations have been updated. The clean_data() no longer takes the params argument. Instead, it takes many other arguments that are described in the function documentation and the package vignettes (#184, Karim-Mane).
Use {cli} package for formatting messages in the package (#196, Karim-Mane).
Translate messages in the package into French using {potools} package (#196, Karim-Mane).
An additional column (row_id) with the row indices where the incorrect date sequences are found is now added to the report made from the check_date_sequence() function (#196, Karim-Mane).
date_guess() and related functions now returns a list of two elements: the converted values and a boolean that informs about the presence of numeric values
that can also be of type Date.
Update the package vignette to account for the improvements made in the different pull requests (#194, Karim-Mane).