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).