cran submission

IndrajeetPatil · Mar 18, 2020 · cfaaef2 · cfaaef2
1 parent bcf8ec2
commit cfaaef2
Show file tree

Hide file tree

Showing 24 changed files with 312 additions and 152 deletions.
diff --git a/.github/CODEOWNERS.txt b/.github/CODEOWNERS.txt
@@ -1,2 +1,2 @@
-# CODEOWNERS for ggstatsplot
+# CODEOWNERS for tidyBF
 .github/CODEOWNERS @IndrajeetPatil
diff --git a/CRAN-RELEASE b/CRAN-RELEASE
@@ -0,0 +1,2 @@
+This package was submitted to CRAN on 2020-03-18.
+Once it is accepted, delete this file and tag the release (commit bcf8ec20b6).
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -11,6 +11,10 @@ Authors@R:
 Maintainer: Indrajeet Patil <patilindrajeet.science@gmail.com>
 Description: Provides helper functions that make it easy to run
     'BayesFactor' package tests on a data which is in a tidy format.
+    Additionally, it provides a more consistent syntax and by default
+    returns a dataframe with rich details. These functions can also return
+    expressions containing results from Bayes Factor tests that can then
+    be displayed on custom plots.
 License: GPL-3 | file LICENSE
 URL: https://indrajeetpatil.github.io/tidyBF/,
     https://github.com/IndrajeetPatil/tidyBF

diff --git a/R/bf_contingency_tab.R b/R/bf_contingency_tab.R
@@ -38,14 +38,12 @@
 #'   \url{https://gist.github.com/richarddmorey/a4cd3a2051f373db917550d67131dba4}.
 #'
 #' @examples
-#'
-#' \donttest{
-#' # ------------------ association tests --------------------------------
-#'
 #' # for reproducibility
 #' set.seed(123)
 #' library(tidyBF)
 #'
+#' # ------------------ association tests --------------------------------
+#'
 #' # to get caption (in favor of null)
 #' bf_contingency_tab(
 #'   data = mtcars,
@@ -54,15 +52,6 @@
 #'   fixed.margin = "cols"
 #' )
 #'
-#' # to get caption (in favor of alternative)
-#' bf_contingency_tab(
-#'   data = mtcars,
-#'   x = am,
-#'   y = cyl,
-#'   fixed.margin = "rows",
-#'   output = "alternative"
-#' )
-#'
 #' # to see results
 #' bf_contingency_tab(
 #'   data = mtcars,
@@ -80,7 +69,6 @@
 #'   x = am,
 #'   prior.concentration = 10
 #' )
-#' }
 #' @export
 
 # function body

diff --git a/R/bf_corr_test.R b/R/bf_corr_test.R
@@ -19,7 +19,6 @@
 #' \code{\link{bf_ttest}}
 #'
 #' @examples
-#' \donttest{
 #' # for reproducibility
 #' set.seed(123)
 #'
@@ -38,7 +37,6 @@
 #'   y = y4,
 #'   bf.prior = 0.8
 #' )
-#' }
 #' @export
 
 # function body

diff --git a/R/helpers_bf_tests.R b/R/helpers_bf_tests.R
@@ -9,14 +9,20 @@
 #' @importFrom dplyr rename select mutate
 #'
 #' @examples
-#' # getting only Bayes Factors
-#' bf_extractor(
+#' set.seed(123)
+#'
+#' # creating a `BayesFactor` object
+#' bf_obj <-
 #'   BayesFactor::anovaBF(
 #'     formula = Sepal.Length ~ Species,
 #'     data = iris,
 #'     progress = FALSE
 #'   )
-#' )
+#'
+#' # extracting Bayes Factors in a dataframe
+#' # *important*: don't enter `1/bf_obj` to extract results for null hypothesis;
+#' # doing so will return wrong results
+#' bf_extractor(bf_obj)
 #' @export
 
 # function body

diff --git a/README.Rmd b/README.Rmd
@@ -15,9 +15,8 @@ knitr::opts_chunk$set(
 )
 ```
 
-# tidyBF: Grouped statistical analysis in a tidy way
+# `tidyBF`: Grouped statistical analysis in a tidy way
 
-[![CRAN_Release_Badge](http://www.r-pkg.org/badges/version-ago/tidyBF)](https://CRAN.R-project.org/package=tidyBF)
 [![packageversion](https://img.shields.io/badge/Package%20version-0.1.0.9000-orange.svg?style=flat-square)](https://github.com/IndrajeetPatil/tidyBF/commits/master)
 [![Daily downloads badge](https://cranlogs.r-pkg.org/badges/last-day/tidyBF?color=blue)](https://CRAN.R-project.org/package=tidyBF)
 [![Weekly downloads badge](https://cranlogs.r-pkg.org/badges/last-week/tidyBF?color=blue)](https://CRAN.R-project.org/package=tidyBF)
@@ -32,11 +31,15 @@ knitr::opts_chunk$set(
 [![minimal R version](https://img.shields.io/badge/R%3E%3D-3.5.0-6666ff.svg)](https://cran.r-project.org/)
 [![Coverage Status](https://img.shields.io/codecov/c/github/IndrajeetPatil/tidyBF/master.svg)](https://codecov.io/github/IndrajeetPatil/tidyBF?branch=master)
 [![Coverage Status](https://coveralls.io/repos/github/IndrajeetPatil/tidyBF/badge.svg?branch=master)](https://coveralls.io/github/IndrajeetPatil/tidyBF?branch=master)
-[![status](https://tinyverse.netlify.com/badge/tidyBF)](https://CRAN.R-project.org/package=tidyBF)
 
 # Overview
 
-`tidyBF` package is a tidy wrapper around `BayesFactor` package.
+`tidyBF` package is a tidy wrapper around `BayesFactor` package that always
+expects the data to be in the tidy format and return a tibble containing Bayes
+Factor values. Additionally, it provides a more consistent syntax and by
+default returns a dataframe with rich details. These functions can also return
+expressions containing results from Bayes Factor tests that can then be
+displayed on custom plots.
 
 # Installation
 
@@ -78,10 +81,12 @@ remotes::install_github(
 Below are few concrete examples of where `tidyBF` wrapper might provide a more
 friendly way to access output from or write functions around `BayesFactor`.
 
-1.  `BayesFactor` is inconsistent with its formula interface. `tidyBF` avoids
-    this as it doesn't provide the formula interface for any of the functions.
+## Syntax consistency
 
-```{r r_formula}
+`BayesFactor` is inconsistent with its formula interface. `tidyBF` avoids this
+as it doesn't provide the formula interface for any of the functions.
+
+```{r formula}
 # setup
 set.seed(123)
 
@@ -105,6 +110,51 @@ bf_ttest(data = mtcars, x = am, y = wt)
 bf_ttest(data = sleep, x = group, y = extra, paired = TRUE)
 ```
 
+## Expressions for plots
+
+Although all functions default to returning a dataframe, you can also use it
+to extract expressions that can be displayed in plots.
+
+```{r expr_plot}
+# setup
+set.seed(123)
+library(ggplot2)
+
+# two-sample t-test results in an expression
+stats_exp <- bf_ttest(mtcars, am, wt, output = "alternative")
+
+# using the expression to display details in a plot
+ggplot(mtcars, aes(as.factor(am), wt)) +
+  geom_boxplot() +
+  labs(subtitle = stats_exp)
+```
+
+## Dataframe with all the details
+
+`BayesFactor` can return the Bayes Factor value corresponding to either
+evidence in favor of the null hypothesis over the alternative hypothesis
+(`BF01`) or in favor of the alternative over the null (`BF10`), depending on how this
+object is called. `tidyBF` on the other hand return both of these values and
+their logarithms.
+
+```{r all_together}
+# `BayesFactor` object
+bf <- BayesFactor::correlationBF(y = iris$Sepal.Length, x = iris$Petal.Length)
+
+# alternative
+bf
+
+# null
+1 / bf
+
+# `tidyBF` output
+bf_corr_test(iris, Sepal.Length, Petal.Length, bf.prior = 0.333)
+```
+
+Note that the log-transformed values are helpful because in case of strong
+effects, the raw Bayes Factor values can be pretty large, but the
+log-transformed values continue to remain easy to work with.
+
 # Code of Conduct
 
 Please note that the `tidyBF` project is released with a [Contributor Code of Conduct](https://contributor-covenant.org/version/2/0/CODE_OF_CONDUCT.html). 

diff --git a/README.md b/README.md
@@ -1,9 +1,8 @@
 
 <!-- README.md is generated from README.Rmd. Please edit that file -->
 
-# tidyBF: Grouped statistical analysis in a tidy way
+# `tidyBF`: Grouped statistical analysis in a tidy way
 
-[![CRAN\_Release\_Badge](http://www.r-pkg.org/badges/version-ago/tidyBF)](https://CRAN.R-project.org/package=tidyBF)
 [![packageversion](https://img.shields.io/badge/Package%20version-0.1.0.9000-orange.svg?style=flat-square)](https://github.com/IndrajeetPatil/tidyBF/commits/master)
 [![Daily downloads
 badge](https://cranlogs.r-pkg.org/badges/last-day/tidyBF?color=blue)](https://CRAN.R-project.org/package=tidyBF)
@@ -22,18 +21,22 @@ Status](https://ci.appveyor.com/api/projects/status/github/IndrajeetPatil/tidyBF
 [![Project Status: Active - The project has reached a stable, usable
 state and is being actively
 developed.](http://www.repostatus.org/badges/latest/active.svg)](http://www.repostatus.org/#active)
-[![Last-changedate](https://img.shields.io/badge/last%20change-2020--03--17-yellowgreen.svg)](https://github.com/IndrajeetPatil/tidyBF/commits/master)
+[![Last-changedate](https://img.shields.io/badge/last%20change-2020--03--18-yellowgreen.svg)](https://github.com/IndrajeetPatil/tidyBF/commits/master)
 [![minimal R
 version](https://img.shields.io/badge/R%3E%3D-3.5.0-6666ff.svg)](https://cran.r-project.org/)
 [![Coverage
 Status](https://img.shields.io/codecov/c/github/IndrajeetPatil/tidyBF/master.svg)](https://codecov.io/github/IndrajeetPatil/tidyBF?branch=master)
 [![Coverage
 Status](https://coveralls.io/repos/github/IndrajeetPatil/tidyBF/badge.svg?branch=master)](https://coveralls.io/github/IndrajeetPatil/tidyBF?branch=master)
-[![status](https://tinyverse.netlify.com/badge/tidyBF)](https://CRAN.R-project.org/package=tidyBF)
 
 # Overview
 
-`tidyBF` package is a tidy wrapper around `BayesFactor` package.
+`tidyBF` package is a tidy wrapper around `BayesFactor` package that
+always expects the data to be in the tidy format and return a tibble
+containing Bayes Factor values. Additionally, it provides a more
+consistent syntax and by default returns a dataframe with rich details.
+These functions can also return expressions containing results from
+Bayes Factor tests that can then be displayed on custom plots.
 
 # Installation
 
@@ -76,10 +79,11 @@ Below are few concrete examples of where `tidyBF` wrapper might provide
 a more friendly way to access output from or write functions around
 `BayesFactor`.
 
-1.  `BayesFactor` is inconsistent with its formula interface. `tidyBF`
-    avoids this.
+## Syntax consistency
 
-<!-- end list -->
+`BayesFactor` is inconsistent with its formula interface. `tidyBF`
+avoids this as it doesn’t provide the formula interface for any of the
+functions.
 
 ``` r
 # setup
@@ -125,6 +129,73 @@ bf_ttest(data = sleep, x = group, y = extra, paired = TRUE)
 #> 1  17.3    1.68e-7 0.0579       2.85      -2.85        1.24       -1.24    0.707
 ```
 
+## Expressions for plots
+
+Although all functions default to returning a dataframe, you can also
+use it to extract expressions that can be displayed in plots.
+
+``` r
+# setup
+set.seed(123)
+library(ggplot2)
+
+# two-sample t-test results in an expression
+stats_exp <- bf_ttest(mtcars, am, wt, output = "alternative")
+
+# using the expression to display details in a plot
+ggplot(mtcars, aes(as.factor(am), wt)) +
+  geom_boxplot() +
+  labs(subtitle = stats_exp)
+```
+
+<img src="man/figures/README-expr_plot-1.png" width="100%" />
+
+## Dataframe with all the details
+
+`BayesFactor` can return the Bayes Factor value corresponding to either
+evidence in favor of the null hypothesis over the alternative hypothesis
+(`BF01`) or in favor of the alternative over the null (`BF10`),
+depending on how this object is called. `tidyBF` on the other hand
+return both of these values and their logarithms.
+
+``` r
+# `BayesFactor` object
+bf <- BayesFactor::correlationBF(y = iris$Sepal.Length, x = iris$Petal.Length)
+
+# alternative
+bf
+#> Bayes factor analysis
+#> --------------
+#> [1] Alt., r=0.333 : 2.136483e+43 ±0%
+#> 
+#> Against denominator:
+#>   Null, rho = 0 
+#> ---
+#> Bayes factor type: BFcorrelation, Jeffreys-beta*
+
+# null
+1 / bf
+#> Bayes factor analysis
+#> --------------
+#> [1] Null, rho = 0 : 4.680589e-44 ±0%
+#> 
+#> Against denominator:
+#>   Alternative, r = 0.333333333333333, rho =/= 0 
+#> ---
+#> Bayes factor type: BFcorrelation, Jeffreys-beta*
+
+# `tidyBF` output
+bf_corr_test(iris, Sepal.Length, Petal.Length, bf.prior = 0.333)
+#> # A tibble: 1 x 8
+#>      bf10 error     bf01 log_e_bf10 log_e_bf01 log_10_bf10 log_10_bf01 bf.prior
+#>     <dbl> <dbl>    <dbl>      <dbl>      <dbl>       <dbl>       <dbl>    <dbl>
+#> 1 2.13e43     0 4.70e-44       99.8      -99.8        43.3       -43.3    0.333
+```
+
+Note that the log-transformed values are helpful because in case of
+strong effects, the raw Bayes Factor values can be pretty large, but the
+log-transformed values continue to remain easy to work with.
+
 # Code of Conduct
 
 Please note that the `tidyBF` project is released with a [Contributor

diff --git a/codemeta.json b/codemeta.json
@@ -5,7 +5,7 @@
   ],
   "@type": "SoftwareSourceCode",
   "identifier": "tidyBF",
-  "description": "Provides helper functions that make it easy to run\n    'BayesFactor' package tests on a data which is in a tidy format.",
+  "description": "Provides helper functions that make it easy to run\n    'BayesFactor' package tests on a data which is in a tidy format.\n    Additionally, it provides a more consistent syntax and by default\n    returns a dataframe with rich details. These functions can also return\n    expressions containing results from Bayes Factor tests that can then\n    be displayed on custom plots.",
   "name": "tidyBF: Tidy Wrapper for 'BayesFactor' Package",
   "codeRepository": "https://github.com/IndrajeetPatil/tidyBF",
   "relatedLink": "https://indrajeetpatil.github.io/tidyBF/",

diff --git a/cran-comments.md b/cran-comments.md
@@ -8,5 +8,6 @@
 0 errors | 0 warnings | 1 note
 
   - This is a new release.
-  - I prefer to keep `LICENSE` file at the top level because this makes it easy
+
+  - I prefer to keep the `LICENSE` file at the top level because this makes it easy
     to read it on GitHub as well.