The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

explore 1.3.0

New features

• add parameter nthread to explain_xgboost(). (#45)
• add interact(). (#47)
• add interactivity as default in explore shiny app. (#47)
• add create_data_abtest().
• add basic color parameter to explore() & abtest() functions.
• add get_color()
• move NA-info in explore() from title to subtitle. (#48)
• add more descriptive infos into explore()subtitle.
• add color parameter for explore(), explore_*(), report()
• add bins parameter to target_explore_num()

Breaking changes

• mix_color() with one color as parameter generates colors from light to dark
• target_explore_num() bar positioning changes from max to mean value

Bug fixes / internal

• fix undefined / not meaningful values in abtest() shiny-app
• rename vignettes explore_*.Rmd to explore-*.Rmd

explore 1.2.0

New features

• add explain_xgboost() (#42)
• add drop_var_by_names() (#43)
• add drop_var_not_numeric() (#43)
• add drop_var_low_variance() (#43)
• add drop_var_no_variance() (#43)
• add drop_var_with_na() (#43)
• add drop_obs_with_na() (#43)
• add drop_obs_if() (#43)
• add mix_color()
• add show_color()
• add create_data_esoteric()

Breaking changes

• create_data_empty() has no longer a parameter seed

Bug fixes / internal

• add check_vec_low_variance() (internal helper function)

explore 1.1.1

• Fix CRAN check warning in running example get_nrow() (#41)

explore 1.1.0

Breaking changes

• explore no longer imports MASS and RandomForest. for explain_logreg() and explain_forest(), you will receive a prompt to install these packages in interactive sessions. (#2 1, @olivroy)

New features

• add explain_forest().
• add predict_target().
• add create_data_newsletter().
• add use_data_beer() and use_data_starwars() functions (#20, #23)
• explore now has a website. (#17, #19, @olivroy)
• abtest() now supports numeric target (t-test).
• abtest_targetpct() with count data (parameter n).
• abtest() and explore() can now run without data (shiny app). If no data are provided, palmerpenguins::penguins is used. (#25)
• New vignettes were added. (#27, #28, #29, #31)
• Documentation and examples were enhanced. (#38, @olivroy, #32, #33, #36, #37)
• create_data_() use_data_*() return data sets as tibble.

Bug fixes / internal

• drop use of fct_explicit_na() (forcats >= 1.0.0) and use linewidth for ggplot2 (>= 3.4.0) (deprecated) (#15, @olivroy)
• explore no longer depends on assertthat, tidyr, and broom. (#16, #21, @olivroy).
• Error messages are now more informative (#18, @olivroy)
• explore now uses testthat 3rd edition and GitHub actions (#16, @olivroy)
• explore has been re-licensed as MIT (#22)

explore 1.0.2 (2023-01-14)

• add_var_random_01() creates variable of type integer
• add target_name & factorise_target parameter to more create_data_*()
• add target1_prob parameter to more create_data_*()
• add checks to create_data_*()
• format variable random_moon
• add abtest()
• remove native pipe in explore_tbl()
• fix error in explore() median if NA values
• add tests for explore() (no error if data contains NA)

explore 1.0.1 (2022-12-20)

• Switch back to %>% in vignettes (compatibility R < 4.1) (#6)
• No hard coded path for markdown-templates
• Add create_data_unfair()
• create_data_app() gains a screen_size argument.

explore 1.0.0 (2022-11-11)

• Dependency DT (>= 0.3.0)
• Improve and use native pipe in README.
• Improve documentation
• Add function create_data_app()
• Add support for integer64
• Bugfix report() >100 variables
• No warning-message in explore_count()
• Redesign explore_tbl()
• Add mean to explore_density() plot
• Add create_data_churn()
• Add add_var_random_moon()
• Vignettes: switch from %>% to |>
• Add create_notebook_explore()

explore 0.9.0 (2022-08-31)

• rename create_x_data() to create_data_x()
• rename add_x_var() to add_var_x()
• extend create_data_*() functions
• extend add_var_*() functions
• explain_tree(): set default minsplit = 20
• explain_tree(): set prior probabilities
• explore() and report(): targetpct as alternative to split parameter
• balance_target(): add parameter seed
• fix variable type for create_data_x()
• using md in roxygen

explore 0.8.0 (2022-01-30)

• all dwh_*() functions are no longer included in {explore} Alternative: source https://github.com/rolkra/dwh
• add create_fake_data()
• add create_random_data()
• add add_random_var()
• add get_var_buckets()
• total_fig_height(): parameters var_name_target, var_name_n
• code styling
• report templates: add var buckets (to plot large number of variables)

explore 0.7.1 (2021-06-04)

• change theme_light() into individual theme() so that set_theme works.
• add rmarkdown to Suggests in DESCRIPTION
• fix URL in DESCRIPTION
• drop LazyData in DESCRIPTION
• format DESCRIPTION

explore 0.7.0 (2021-01-21)

• explain_tree() gains a weights parameter.
• flip no/yes label in rpart.plot
• fix minsplit for count-data
• add weight_target()
• add plot_legend_targetpct()
• add legend in targetpct rmarkdown-template
• add unit testing (testthat)

explore 0.6.2 (2020-10-14)

• Bugfix explore_bar(): NA in plot
• explore_count(): convert target into factor
• explore_count(): add default title (cat name)
• explore_count(): add parameter numeric, max_cat, max_target_cat
• explain_tree(): convert character variables into factors (count data)
• explain_tree(): parameter out (“plot” | “model”)
• explain_logreg(): parameter out (“tibble” | “model”)
• vignette("explore_titanic"): change to tibble
• vignette("explore_mtcars"): add explanations
• change theme_minimal to theme_light
• dwh_fastload(): add parameters overwrite and append
• update README.md

explore 0.6.1 (2020-09-04)

• Fix Github URL
• new Vignette vignette("explore_penguins")
• new Vignette vignette("explore_titanic") (count data)
• explore_count(): plot count() output
• add default parameter n for count data: explore(), explore_all(), explore_tbl(), explain_tree(), report(), describe(), describe_cat(), describe_num(), describe_tbl(), total_fig_height()
• explore_tree(): default value for minsplit = 10% of obs
• explore_cor(): use geom_point() for small datasets
• explore_shiny(): use browseURL() with parameter browser=NULL
• describe_tbl(): add observations containing NA
• guess_cat_num(): parameter description (optional)
• count_pct(): no renaming of variables.

explore 0.5.5 (2020-04-06)

Maintenance update:

• fix breaking changes tibble 3.0.0

explore 0.5.4 (2020-02-09)

Maintenance update:

• fix param ... in description (PR#16223, see https://bugs.r-project.org/show_bug.cgi?id=16223)

explore 0.5.3 (2020-01-17)

• explore_bar(): add parameter numeric
• describe_all() returns a tibble
• describe_all(): column ‘variable’ is character (not factor)
• report() split = TRUE as default
• add rescale01()
• add parameter rescale01 to clean_var()
• add function count_pct()
• add out='tibble' to describe_cat()
• add function explore_targetpct()

explore 0.5.2 (2019-11-22)

• split source-code file into multiple files
• format_num_auto() without brackets
• treat Date variables as cat
• report() fix automatic file extension .html
• add simplify_text()
• add parameter simplify_text to clean_var()
• fix link in README.md

explore 0.5.1 (2019-10-08)

Prepare for new dplyr 0.8.4 (#2, @romainfrancois)

Bug Fixes

• prepare explore_tbl() for dplyr 0.8.4
• describe_num() with default digits=6
• describe_cat() bugfix variable with all NA
• describe_all() bugfix variable with all NA
• explain_tree() bugfix dataframe with 0 rows
• improve speed describe() text output (RMarkdown)
• explore() now checks if data is a data.frame

explore 0.5.0 (2019-09-19)

Interactive data exploration now accept categorical and numerical targets (next to a binary target).

• explain_tree(): target can be bin/num/cat
• explain_tree(): add parameter max_target_cat
• explore_shiny(): target can be bin/num/cat
• add function format_num_auto()
• total_fig_height() replaces the now deprecated get_nrow().
• add parameter title to explore_cor()
• add support for POSIXct in describe()
• improved handling of dataframes with no observations
• add parameter title to explore_density()
• add parameter nvar to total_fig_height()
• update README.md
• update Vignettes
• add NEWS.md
• add hex sticker

explore 0.4.4 (2019-08-27)

Many functions now accept categorical and numerical targets (next to a binary target). If you want to force which geom is used for visualisation, you can use explore_bar() and explore_density(). New function explore_tbl() to visualise a dataframe/table (type of variables, number of NA, …)

• add function explore_bar()
• explore_density() now using correct tidy eval, target cat > 2 possible
• target_explore_cat() now using correct tidy eval
• target_explore_num() now using correct tidy eval
• add plot_var_info() - plots a info-text to a variable as ggplot obj.
• plot_var_info() used in explore/explore_all if
• plot_var_info() used if explore empty data
• add parameter max_cat in explore_bar(), explore_density() and explain_tree()
• add explore_tbl()
• drop explore_cat() & explore_num()
• rename template_report_target_den.html > _split.html
• intelligent placing of labels in plots
• add info window “generating report …” in explore_shiny()
• format_num() -> format_num_kMB(), format_num_space()
• format_target() -> if numeric split 0/1 by mean
• report() -> default .html file extension
• consistency showing NA info in explore-title
• parameter split: default = FALSE
• allow numeric (num) target in explore_all & report
• describe_tbl() -> fix target if not bin
• describe(): change out=“vector” to out=“list”

explore 0.4.3 (2019-06-17)

• fix parameter in explore(): auto_scale, na
• fix number of NA in explore() (move code before auto_scale)
• explore_density() with target: drop plot title “propensity by”
• explore_shiny(): use output_dir / tempdir()
• change naming “attribute” to “variable” (consistent)

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.