Purpose
The goal of tidytidbits is to complement the environment provided by
the tidyverse packages with a set of higher-level functions within the
scope of data analysis and plotting.
Installation
Install from CRAN:
install.packages("tidytidbits")
Or install the latest git version from bitbucket:
devtools::install_bitbucket("mwiesweg/tidytidbits")
Components
The following areas are covered:
- methods designed for use in a dplyr pipeline
- formatting columns: format_numbers_at, format_p_values_at, and the
corresponding generic tools as_formatted_number, as_formatted_p_value,
as_percentage_label
- adding data analysis results as new columns: count_by,
add_prop_test
- cross tabulation in a pipeline: cross_tabulate
- lumping rows of a data frame by a count: lump_rows, and the
corresponding generic tool lump
- appending and prepepnding in a pipe with different semantics:
append_object, prepend_object
- lower-level pipeline tools: add_summary, add_summary_by, execute_if,
execute_in_pipeline, interlude
- convenience methos to rename and reorder a factor:
rename_reorder_factor, rename_factor, order_factor_by
- vectorised and non-vectorised methods implementing usefuls notions
about NA and logical values: all_or_all_na, any_or_all_na, are_true,
equal_including_na, falsy, invalid, true_or_na, truthy, valid
- vectorised methods dealing with data frame columns potentially
containing NA values: first_non_nas, first_not, first_not_na,
first_which_not_na, which_non_na
- a vectorised lookup method for use with dictionary-like vectors:
lookup and its type-specific variants
- some accessors for use with purrr’s pluck or the additionally
provided pluck_vector: name_contains, named, value_contains
- generic tools: invert_value_and_names, str_locate_match
- dealing with sequential duplicates: sequential_duplicates,
replace_sequential_duplicates
- for plotting:
- conveniently saving a plot from a pipe: save_pdf, save_png
- convenience methods for safe directory generation: prepare_path,
prepare_directory
- convenience method giving frequently used paper sizes: dinA_format,
dinA_height, dinA_width
- convenience method to create a named palette: named_palette and for
creating a qualitative palette with many entries: iwanthue_palette
- a notion of having a method which provides its local variables for
reuse in other methods: local_variables and the corresponding
source_variables
- a python-style tuple-assignment implementation
g(a, b) %=% function_returning_vector_of_two_elements()
- some low-level methods utilizing rlang features: eval_unquoted,
expression_list, print_deparsed, quosure_list, symbol_as_quosure,
symbol_string_list, syntactically_safe