- bug fix in qroc - use
`geom_path`

instead of`geom_line`

- extend return section of
`confusion_matrix`

to fix CRAN NOTE

- Unit testing is done in base R, moved away from testthat.
- Updating roxygen dependencies to correctly document the package.
- Extend documentation and testing of
`set_diff`

- The default value for
`conf_int`

in`qacf`

has been modified from`0.05`

to`1 - getOption("qwraps2_alpha", 0.05)`

`ggplot2_extract_legend`

is now and S3 method`qable`

is now and S3 method and has gained a new argument`kable_args`

to make it clear what values are used by`qable`

and which are passed to`knitr::kable`

(#84)`summary_table`

has gained two arguments,`qable_args`

and`kable_args`

to help with passing through arguments and being able to use the features of knitr::kable (#84)- vignettes have been renamed to have a consistent “qwraps2-
” structure. `summary_table`

has it’s own vignette.

`mean_se`

to provide the mean and standard error of the mean in the same format as the`mean_sd`

function.Add Precision-Recall curve,

`qprc`

, to complement ROC,`qroc`

.Add quick “restricted mean survival time,”

`qrmst`

method.Binary Operators:

`%s%`

for “adding” strings together, e.g.,`"A" %s% "B" %s% "C"`

is equivalent to`paste0("A", "B", "C")`

add

`append`

argument to`frmt`

`qable`

has been modified to return a character matrix with the S3 class`qwraps2_qable`

. A`print`

method has been added as a wrapper to`knitr::kable`

to generate the table with the wanted markup. This change will allow end users to modify the matrix generated by`qable`

and print the markup results with ease. This will be particularly useful when using`summary_table`

and added bespoke p-values to the rows in a “table 1.”`spin_comments`

returns`TRUE`

or`FALSE`

with the notes on the matching/mismatched comments as an attribute of the return object.Extend the S3 methods provided for

`qroc_build_data_frame`

**Potentially breaking Change**`confusion_matrix`

has been undergone a major refactor. If nothing else, the return object is now a data.frame with more output than before. The inputs are hopefully easier for end users as well. This change includes the addition of the function`auc`

and major refactor of`qroc`

and`qprc`

.

`summary_table_042`

and`tab_summary`

, the tidyverse versions of summary table that had been deprecated have been removed from the package. (#96)`qroc_build_data_frame`

as been deprecated in favor of`auc`

- Correct error in summary table cbind (#104)

- Correct error in output of bootstrapped results for confusion matrices (#102)

`set_diff`

`spin_comments`

- a tool to identify opening and closing pairs of comments in a .R file expected to go through knitr::spin. Built to help resolve the “comments must be put in pair of start and end delimiters” error that can be thrown by knitr::spin.`mtcars2`

is a fully documented data set provided as part of the qwraps2 package.`summary_table`

has a`by`

argument and supports summaries by more than one variable.The example data set

`mtcars2`

is now exported as part of the package. The construction of this data set is documented in a new vignette,`vignette("qwraps2-data-sets")`

A data set

`pefr`

for peak expiratory flow rate data from Martin and Altman- has been added to the package. Documentation of the data set is
found in the
`vignette("qwraps2-data-sets")`

- has been added to the package. Documentation of the data set is
found in the
Examples, and extended details, for graphics have been added to new vignette

`vignette("qwraps2-graphics")`

- Scoping issue (#85) in the confusion matrix call
- Long object names passed to
`summary_table`

no longer generate an error (#90)

confusion matrix returns additional statistics (#87)

- F1
- MCC

`...`

passes values arguments to`stats::predict`

via`qroc`

and`qroc_build_data_frame`

reduced the use of non-base R, specifically tidyverse and magrittr, within the base code of qwraps2. Focus on base R methods instead. This will, hopefully, improve long term stability of the package and reduce warnings and error generated by an ever changing tidyverse api. This includes a

`by`

argument (#100), and other notes related to #96, #98, and #71.deprecated the transform argument in

`mean_ci`

– to easy to have bad results come out of that option. Think about the delta method.

If the a variable in a data.frame has an attributed called “label” then that label will be used in the rgroup for

`qsummary`

and`summary_table`

, (#74)`backtick`

is provided as a work around for a`knitr::spin`

issue. The function wraps text in backticks so that backticks do not need to be explicitly used in a .R file expected to be spun to .Rmd.

Comment out a test due to the change in random number generating for
`sample`

. (#72)

`rbind.qwraps2_summary_table`

(#55, #56)- formatting options for package names and urls to CRAN, gitlab, and github. (#68)

`cbind.qwrap2_summary_table`

check attributes of objects to be combined together. (#56)- Added example to vignette to address the use
`tab_summary`

with more than one variable, e.g., in a`sapply`

call. (#53) `qsummary`

will return the first (min) and last (max) for variables where`inherit(var, "Date")`

is`TRUE`

.

`tab_summary`

has been deprecated, use`qsummary`

instead. A warning message will be given to the user at this time.`tab_summary`

will be removed completely in a later version.`create_pkg`

has been deprecated and removed from the package. The function was going to need a massive rewrite to deal with changes to devtools and usethis. Effectively, the function would no longer work and this deprecation is just being pragmatic.

Added

`lazyload_cache_dir`

and`lazyload_cache_labels`

for (lazy)loading of cached chunks. Very useful for loading cached`knitr`

code chunks into an interactive R session.Added

`traprule`

for a quick way to integrate a numeric vectorAdded

`create_pkg`

: a wrapper about`devtools::create`

with more defaults that I prefer.Added a

`file_check`

(#50) to check for read access and optionally md5sum of files.Added a

`pkg_check`

(#51) to check if a list of packages are available.Added functions for geometric mean, variance, and standard deviation. (#52) See the summary-statistics vignette for details.

Extended the documentation for

`logit`

and`invlogit`

to reference bases R methods.Added to the examples for

`summary_table`

showing how to add a caption to a LaTeX table (#39)step ribbon used in

`qkmplot`

(#45)

Factor levels as characters in

`summary_table`

(#48)`qkmplot`

works for intercept only models (#38)

`frmtci.data.frame`

method added.- for the
`tab_summary`

methods to new arguments:`n_perc_args`

and`envir`

. The former gives the end user control over the formatting options passed to`n_perc`

and the latter controls the`environment`

associated with the generated formulae.

`frmtci.matrix`

using the`est`

arg correctly

- extended documentation and examples for
`summary_table`

`qacf`

plot (#11)`ll`

an improved version of`ls()`

`rtitle`

option added to`qable`

(#35)

`perc_n`

reports the strings in the correct order.

- Extended documentation for
`qable`

- Minor edits and extensions to the summary-statistics vignette.

`summary_table`

and`tab_summary`

added (#33)- New vignette
`vignette("summary-statistics, package = "qwraps2")`

- boot strapped statistics in confusion matrices

`logit`

and`invlogit`

functions added. Using Rcpp so they are fast.

`format`

is correctly passed from`qwraps2::qable`

to`knitr::kable`

’s`format`

argument.

This is a long over due version bump. Many new features and bug fixes have been made since 0.1.0.

`confusion_matrix`

produces the correct confusion matrix. Prior version could transpose the result.`gmean_sd`

displays the correct geometric standard deviation

`mean_sd`

and`gmean_sd`

have better`show_n`

handling.

`print.confusion_matrix`

provides more detail to make it clear which variable is the Truth and which is the Prediction.`ggplot2_extract_legend`

added (#6)`show_n`

functionality for`median_iqr`

implemented. (closes #24)`frmtci`

added (#32)`mean_ci`

added

`confusion_matrix.formula`

no longer uses named columns. The function only worked when called from`confusion_matrix.default`

. Fix allows the function to be used as intended.

`gmean_sd`

returns the geometric mean and standard deviation

- The
`confusion_matrix`

method has been modified. Please read the documentation. If you used version 0.1.2 these changes will affect your work. The`confusion_matrix`

is now an S3 generic.

- methods for finding the sensitivity and specificity of confusion matrices (contingency tables) added.
`median_iqr`

returns a formatted string with the median and IQR of a data vector.`perc_n`

similar to the`n_perc`

but returns a string of the form`p% (n = N)`

.`n_perc`

has option to suppress the percent symbol`n_perc0`

added in 0.1.1.9004 sets defaults which may be preferred for tables.`qkmplot`

for plotting Kaplan Meier curves`qable`

creates`knitr::kable`

tables with row groups and names similar to`hmisc::latex`

or`htmlTable::htmlTable`

.

Pushed to CRAN on 20 April 2015.

`extract_fstat`

extracts the F-statistic from`lm`

`extract_fpvalue`

extracts and formats the omnibus F-test p-value from`lm`

- P-value formatting for Obstetrics & Gynecology

- none

- Set R dependence to 3.0.2 to match the dependence noted for dplyr
- Default setting for the qwraps2_frmtp_digits is 4 instead of 2.
- Spelling and grammar fixes in documentation (#21)

`mean_sd`

passes the`digits`

option to`frmt`

correctly (#20)

Initial release.