tidymodels · topepo · May 19, 2020 · May 7, 2020 · May 7, 2020 · May 7, 2020
diff --git a/NEWS.md b/NEWS.md
@@ -1,5 +1,7 @@
 # rsample (development version)
 
+* Lower threshold for pooling strata to 10% (from 15%) (#149).
+
 # `rsample` 0.0.6
 
 * Added `validation_set()` for making a single resample.

diff --git a/R/boot.R b/R/boot.R
@@ -12,8 +12,9 @@
 #'  bootstrap results.
 #' The `strata` argument is based on a similar argument in the random forest
 #'  package were the bootstrap samples are conducted *within the stratification
-#'  variable*. The can help ensure that the number of data points in the
+#'  variable*. This can help ensure that the number of data points in the
 #'  bootstrap sample is equivalent to the proportions in the original data set.
+#'  (Strata below 10% of the total are pooled together.)
 #' @inheritParams vfold_cv
 #' @param times The number of bootstrap samples.
 #' @param strata A variable that is used to conduct stratified sampling. When

diff --git a/R/initial_split.R b/R/initial_split.R
@@ -5,9 +5,9 @@
 #'  _first_ `prop` samples for training, instead of a random selection.
 #'  `training` and `testing` are used to extract the resulting data.
 #' @details The `strata` argument causes the random sampling to be conducted
-#'  *within the stratification variable*. The can help ensure that the number of
-#'  data points in the training data is equivalent to the proportions in the
-#'  original data set.
+#'  *within the stratification variable*. This can help ensure that the number
+#'  of data points in the training data is equivalent to the proportions in the
+#'  original data set. (Strata below 10% of the total are pooled together.)
 #' @inheritParams vfold_cv
 #' @param prop The proportion of data to be retained for modeling/analysis.
 #' @param strata A variable that is used to conduct stratified sampling to

diff --git a/R/make_strata.R b/R/make_strata.R
@@ -66,7 +66,7 @@
 #' quantile(x6, probs = (0:10)/10)
 #' table(make_strata(x6, breaks = 10))
 #' @export
-make_strata <- function(x, breaks = 4, nunique = 5, pool = .15, depth = 20) {
+make_strata <- function(x, breaks = 4, nunique = 5, pool = .1, depth = 20) {
   num_vals <- unique(x)
   n <- length(x)
   num_miss <- sum(is.na(x))

diff --git a/R/mc.R b/R/mc.R
@@ -4,9 +4,9 @@
 #'  replacement) of the original data set to be used for analysis. All other
 #'  data points are added to the assessment set.
 #' @details The `strata` argument causes the random sampling to be conducted
-#'  *within the stratification variable*. The can help ensure that the number of
+#'  *within the stratification variable*. This can help ensure that the number of
 #'  data points in the analysis data is equivalent to the proportions in the
-#'  original data set.
+#'  original data set. (Strata below 10% of the total are pooled together.)
 #' @inheritParams vfold_cv
 #' @param prop The proportion of data to be retained for modeling/analysis.
 #' @param times The number of times to repeat the sampling.

diff --git a/R/validation_split.R b/R/validation_split.R
@@ -6,7 +6,7 @@
 #' @details The `strata` argument causes the random sampling to be conducted
 #'  *within the stratification variable*. This can help ensure that the number of
 #'  data points in the analysis data is equivalent to the proportions in the
-#'  original data set.
+#'  original data set. (Strata below 10% of the total are pooled together.)
 #' @inheritParams vfold_cv
 #' @param prop The proportion of data to be retained for modeling/analysis.
 #' @param strata A variable that is used to conduct stratified sampling to

diff --git a/R/vfold.R b/R/vfold.R
@@ -7,9 +7,9 @@
 #'  to V.
 #' @details
 #' The `strata` argument causes the random sampling to be conducted *within
-#'  the stratification variable*. The can help ensure that the number of data
+#'  the stratification variable*. This can help ensure that the number of data
 #'  points in the analysis data is equivalent to the proportions in the original
-#'  data set.
+#'  data set. (Strata below 10% of the total are pooled together.)
 #' When more than one repeat is requested, the basic V-fold cross-validation
 #'  is conducted each time. For example, if three repeats are used with `v =
 #'  10`, there are a total of 30 splits which as three groups of 10 that are

diff --git a/README.Rmd b/README.Rmd
@@ -13,7 +13,7 @@ knitr::opts_chunk$set(
 )
 ```
 
-# rsample
+# rsample <a href='https://tidymodels.github.io/rsample/'><img src='man/figures/logo.png' align="right" height="139" /></a>
 
 <!-- badges: start -->
 [![R build status](https://github.com/tidymodels/rsample/workflows/R-CMD-check/badge.svg)](https://github.com/tidymodels/rsample/actions)
@@ -23,7 +23,10 @@ knitr::opts_chunk$set(
 ![](https://img.shields.io/badge/lifecycle-maturing-blue.svg)
 <!-- badges: end -->
 
-`rsample` contains a set of functions that can create different types of resamples and corresponding classes for their analysis. 
+
+## Overview
+
+`rsample` contains a set of functions to create different types of resamples and corresponding classes for their analysis. 
 The goal is to have a modular set of methods that can be used across different R packages for:
 
  * traditional resampling techniques for estimating the sampling distribution of a statistic and
@@ -57,8 +60,11 @@ as.numeric(lobstr::obj_size(boots)/lobstr::obj_size(LetterRecognition))
 #> [1] 2.528326
 ```
 
+<sup>Created on 2020-05-07 by the [reprex package](https://reprex.tidyverse.org) (v0.3.0)</sup>
+
 The memory usage for 50 bootstrap samples is less than 3-fold more than the original data set. 
 
+
 ## Installation
 
 To install it, use:
@@ -84,7 +90,7 @@ We welcome contributions of all types!
 
 If you have never made a pull request to an R package before, `rsample` is an excellent place to start. Find an [issue](https://github.com/tidymodels/rsample/issues/) with the **Beginner Friendly** tag and comment that you'd like to take it on and we'll help you get started.
 
-We encourage typo corrections, bug reports, bug fixes and feature requests. Feedback on the clarity of the documentation is especially valuable.
+We encourage typo corrections, bug reports, bug fixes, and feature requests. Feedback on the clarity of the documentation is especially valuable.
 
 
 ## Code of Conduct

diff --git a/README.md b/README.md
@@ -1,38 +1,28 @@
 
-# rsample
+# rsample <a href='https://tidymodels.github.io/rsample/'><img src='man/figures/logo.png' align="right" height="139" /></a>
 
 <!-- badges: start -->
-
-[![R build
-status](https://github.com/tidymodels/rsample/workflows/R-CMD-check/badge.svg)](https://github.com/tidymodels/rsample/actions)
-[![Codecov test
-coverage](https://codecov.io/gh/tidymodels/rsample/branch/master/graph/badge.svg)](https://codecov.io/gh/tidymodels/rsample?branch=master)
-[![CRAN\_Status\_Badge](http://www.r-pkg.org/badges/version/rsample)](https://cran.r-project.org/package=rsample)
+[![R build status](https://github.com/tidymodels/rsample/workflows/R-CMD-check/badge.svg)](https://github.com/tidymodels/rsample/actions)
+[![Codecov test coverage](https://codecov.io/gh/tidymodels/rsample/branch/master/graph/badge.svg)](https://codecov.io/gh/tidymodels/rsample?branch=master)
+[![CRAN_Status_Badge](http://www.r-pkg.org/badges/version/rsample)](https://cran.r-project.org/package=rsample)
 [![Downloads](http://cranlogs.r-pkg.org/badges/rsample)](https://cran.r-project.org/package=rsample)
 ![](https://img.shields.io/badge/lifecycle-maturing-blue.svg)
 <!-- badges: end -->
 
-`rsample` contains a set of functions that can create different types of
-resamples and corresponding classes for their analysis. The goal is to
-have a modular set of methods that can be used across different R
-packages for:
-
-  - traditional resampling techniques for estimating the sampling
-    distribution of a statistic and
-  - estimating model performance using a holdout set
 
-The scope of `rsample` is to provide the basic building blocks for
-creating and analyzing resamples of a data set but does not include code
-for modeling or calculating statistics. The “Working with Resample Sets”
-vignette gives demonstrations of how `rsample` tools can be used.
+## Overview
 
-Note that resampled data sets created by `rsample` are directly
-accessible in a resampling object but do not contain much overhead in
-memory. Since the original data is not modified, R does not make an
-automatic copy.
+`rsample` contains a set of functions to create different types of resamples and corresponding classes for their analysis. 
+The goal is to have a modular set of methods that can be used across different R packages for:
+
+ * traditional resampling techniques for estimating the sampling distribution of a statistic and
+ * estimating model performance using a holdout set
+
+The scope of `rsample` is to provide the basic building blocks for creating and analyzing resamples of a data set but does not include code for modeling or calculating statistics. The "Working with Resample Sets" vignette gives demonstrations of how `rsample` tools can be used.  
 
-For example, creating 50 bootstraps of a data set does not create an
-object that is 50-fold larger in memory:
+Note that resampled data sets created by `rsample` are directly accessible in a resampling object but do not contain much overhead in memory. Since the original data is not modified, R does not make an automatic copy. 
+
+For example, creating 50 bootstraps of a data set does not create an object that is 50-fold larger in memory:
 
 ``` r
 library(rsample)
@@ -56,9 +46,12 @@ as.numeric(lobstr::obj_size(boots)/lobstr::obj_size(LetterRecognition))
 #> [1] 2.528326
 ```
 
+<sup>Created on 2020-05-07 by the [reprex package](https://reprex.tidyverse.org) (v0.3.0)</sup>
+
 The memory usage for 50 bootstrap samples is less than 3-fold more than
 the original data set.
 
+
 ## Installation
 
 To install it, use:
@@ -80,21 +73,12 @@ install_dev("rsample")
 
 ## Contributing
 
-We welcome contributions of all types\!
+We welcome contributions of all types!
 
-If you have never made a pull request to an R package before, `rsample`
-is an excellent place to start. Find an
-[issue](https://github.com/tidymodels/rsample/issues/) with the
-**Beginner Friendly** tag and comment that you’d like to take it on and
-we’ll help you get started.
+If you have never made a pull request to an R package before, `rsample` is an excellent place to start. Find an [issue](https://github.com/tidymodels/rsample/issues/) with the **Beginner Friendly** tag and comment that you'd like to take it on and we'll help you get started.
 
-We encourage typo corrections, bug reports, bug fixes and feature
-requests. Feedback on the clarity of the documentation is especially
-valuable.
+We encourage typo corrections, bug reports, bug fixes, and feature requests. Feedback on the clarity of the documentation is especially valuable.
 
 ## Code of Conduct
 
-Please note that the rsample project is released with a [Contributor
-Code of
-Conduct](https://tidymodels.github.io/rsample/CODE_OF_CONDUCT.html). By
-contributing to this project, you agree to abide by its terms.
+Please note that the rsample project is released with a [Contributor Code of Conduct](https://tidymodels.github.io/rsample/CODE_OF_CONDUCT.html). By contributing to this project, you agree to abide by its terms.
diff --git a/man/bootstraps.Rd b/man/bootstraps.Rd
diff --git a/man/initial_split.Rd b/man/initial_split.Rd
diff --git a/man/make_strata.Rd b/man/make_strata.Rd
diff --git a/man/mc_cv.Rd b/man/mc_cv.Rd
diff --git a/man/validation_split.Rd b/man/validation_split.Rd
diff --git a/man/vfold_cv.Rd b/man/vfold_cv.Rd
diff --git a/tests/testthat/test_strata.R b/tests/testthat/test_strata.R
@@ -23,7 +23,7 @@ test_that('simple character', {
 })
 
 test_that('bad data', {
-  x3 <- factor(rep(LETTERS[1:10], each = 50))
+  x3 <- factor(rep(LETTERS[1:15], each = 50))
   expect_warning(make_strata(x3))
   expect_warning(make_strata(mtcars$mpg))
 })