NEWS.md
In mde: Missing Data Explorer

title: 'Missing Data Explorer' author: "Nelson Gonzabato" date: "r Sys.Date()" output: html_document

mde 0.3.2

There is now a function recode_as_value for general recoding. See #41.
Fixed all_of warnings in dplyr select and across verbs.
Tests now use edition 3 of testthat. See https://github.com/r-lib/testthat/issues/1471 and https://github.com/Nelson-Gon/manymodelr/pull/22
Vignettes no longer include URLs to avoid future issues with broken links on CRAN.
Exclusion or inclusion in na_summaryvia regular expressions is now possible via regex_kind and providing a pattern_type and pattern.
exclusion is now done entirely in base R instead of using dplyr::select.
get_na_means and percent_missing now support POSIXct.
na_counts and percent_na are new vector focused functions to allow getting na_counts and percent missingness for objects of classes like POSIXct.
include a new argument to reset rownames in na_summary. Fixes #33.

Fixed a bug that made exclude_cols result in non-exclusion in grouped na.summary outputs.
na_summary now includes a round_to parameter to control the number of digits to round to. Defaults to options("digits") in the user's settings.
drop_na_at no longer uses a tibble to avoid warnings with tibble row.names.
Tests in dict_recode were extended to ensure that we get what is expected particularly with respect to the use_func argument.
dict_recode now supports the use of different NA recoding functions. These can be provided via the argument use_func.
dict_recode is a new function to allow recoding with a pattern-values pair.
drop_na_if now supports keep certain columns when grouping. It also now supports dropping only at certain columns.
drop_na_if now supports dropping groups that meet a set missing data threshold.
recode_as_na_str updated to convert factors to character.

drop_row_if arguments were changed. See docs for the new arguments. This was done to allow new functionality that enables dropping rows based on counts.
Sorting strings in na_summary contained a minor bug that has now been fixed.
recode_as_na_str is a new function to recode_as_na based on string matching.
recode_as_na_if is a new function that allows deliberate addition of NAs to a column if it meets a pre-defined condition. This is essentially the reverse of drop_na_if.
drop_row_if is a new function similar to drop_na_if but for rows instead of columns.
percent_missing can now work independently for factors, characters, or numeric data. It previously was limited to data.frame objects.
Fixed issues with drop_na_at that resulted in failing data rebinding.
get_na_counts now includes an optional exclude_cols argument to exclude columns from the analysis.
na_summary has been fixed to allow exclusion of columns from the analysis.
all_na is a new low level function to perform missingness boolean checks for columns(vectors).
Added a new function column_based_recoode that aims to allow NA recoding based on other columns.

Additions

na_summary which provides a very quick overview of missingness. It also supports grouped summaries.
drop_na_if allows easy dropping of columns where all values are missing.
custom_na_recode allows replacing missing values with common values such as mean, min, max, sd.

Major changes

In percent_missing, the argument grouped was dropped in favour of simply providing a grouping_cols vector.
In recode_as_na , subset_df was dropped. It now simply accepts an optional subset_cols argument. The argument tidy was also dropped.One can simply provide an optional pattern_type and pattern.
Similar changes were made for recode_na_as as above.
The argument x was changed to df in drop_na_at.

Available functions

Key changes

Now supports tidy selection and exploration at specific columns
percent_missing supports grouping and exclusion of certain columns. The use of decimals was dropped.
drop_na_if allows exclusion of columns.