![[BACK]](/icons/back.gif){kind=icon}
Up to [cvs.NetBSD.org] / pkgsrc / misc / R-Hmisc
Request diff between arbitrary revisions
Keyword substitution: kv
Default branch: MAIN
(misc/R-Hmisc) Updated 5.1.0 to 5.1.1
Changes in version 5.1-1 (2023-09-11)
* meltData: generalized to allow melting on either left or right
side of formula, added ... arguments passed to label(), made
variable a factor to keep original order
* formatCons: fixed bug where lang != 'latex' should have been lang
== 'plain' which made msdsize to be ignored for html output
* spikecomp: made more reliable by not assuming as much about
pretty() boundaries
* redun: added rank and qrank options
* added print and plot methods for princmp
* redun: added ability to use numeric matrix instead of formula
* ggplot.transcan: fixed bug when adata is empty
* ggplot.transace: new function
* transace: added trantab result, added as.matrix(x) so can run on
data frames of all numeric variables; added automatic setting of
binary, category, etc. if x is a data frame/table;
non-downward-compatible change: made returned object of class
transace with different format; remove pr and pl arguments from
transace. Changed first argument to a formula and required
monotone(), categorical(), linear() to wrap variables in the
formula instead of specifying with separate arguments
* runParallel: put in NAMESPACE
* qcrypt: new function for encrypting and decrypting data with a
safe workflow
* aregImpute: added constraint argument to allow general
constraints on values imputed with predictive mean matching
(misc/R-Hmisc) Updated 4.8.0 to 5.1.0
Changes in version 5.1-0 (2023-05-04)
* fit.mult.impute: added robust, cluster, robmethod arguments to
incorporate rms::robcov to do robust sandwich covariance
estimation with multiple imputation
* testCharDateTime: new function
* added totxt function in markupSpecs$html to convert html to plain text
* describe.vector: fixed for mChoice variables with plain output, fixed NA dection for mChoice in general; added shortmChoice option, now defaulting to printing mChoice summary using numeric codes, and printing levels once up front; changed back to old behavior of not printing extremes if frequency table is printed; changed printing of lowest and highest 5 values to 6 significant digits if numeric; added argument rmhtml to strip html from variable labels
* mChoice: changed result to NA instead of '' if no choices are made and add.note=FALSE; fixed label; improved print.summary.mChoice formatting
* htmlVerbatim: removed blank line before </pre>
* htmltabv: new function to make simple html table of a series of verbatim outputs from print() of objects
* summary.mChoice: added short= argument
* print.summary.mChoice: if options(prType='html') is in effect, calls htmltabv to compose output in a space-savings 2-column html table, and renders entire output in html
* describe: retrieved format* attribute of variables
* vlab: fixed bug
* dotchartp: fixed bug when sort was logical resulting in o=NULL; changes in R made x[NULL,] quit working
* vlab, hlab, hlabs: improved logic and also look in global environment
* spikecomp: added new options that facilitate sparklines
* html.data.frame: added new argument disableq to disable Quarto table style
* describe: for html use disableq for html.data.frame
* added gt, sparkline, jsonlite packages to SUGGESTS
* added which= argument to print.describe
* html_describe_cat, html_describe_con: new internal functions to print gt tables when which= is given to print.describe
* describe: for date/time variables use lowest observed value for leftmost point of first bin, instead of pretty()
* describe: compute min, max, mean no. of characters (after trimming white space) for character variables that are not tabulated because of too many levels; used by html_describe_cat; also added mode and its count for this situation
* htmlwidgets: added to SUGGESTS
* html.describe: changed html to Hmisc::html to avoid conflicts with gt
* describe: added trans argument to transform x for the purpose of making spike histograms
* spikecomp: generalized, describe.vector now calls it, changed how addggLayers calls it
* fit.mult.impute: added method= including stacking to facilitate likelihood ratio tests with rms::processMI
* histboxp, dhistboxp: fixed bugs where range() should have been diff(range())
* fit.mult.impute: for stacked ols fits correct residual d.f.
Changes in version 5.0-1 (2023-03-06)
* rendHTML: when knitr is not running and object is not html, assumes it is RMarkdown and calls rmarkdown::render to render it to html, then displays in viewer
* princmp: new function to help in interpreting principal components and sparse principal components
* getabd: new function to fetch datasets from The Analysis of Biological Data
* runParallel: new function to make the parallel package easy to use and to improve structure of combined final result
* hashCheck: new function to run digest::digest on a series of argument to create a hash, fetch an existing result file which contains the hash of the input objects the last time an analysis was run, and to return the results stored in the file (an .rds file) if the hashes match, or NULL otherwise
* runifChanged: new function to re-run code if an input changed, as judged by hashChech
* hlab: new function to retrieve plotting-formatted variable label from a current dataset or from the object created by extractlabs (LabelsUnits, which takes priority)
* hlabs: new function to call ggplot2 labs() after running variable names through hlab()
* vlab: new function like hlab but returns text string form of label/units
* extractlabs: new function: for one or more data frames/tables saves a data table of all variables that had a non-blank label or units attribute.
* nCoincident: new function to count the number of x,y pairs that are likely to be coincident, i.e., hidden from view in a scatterplot
* meltData: new function to take a formula and melt a data frame/table so that all right-hand-side formula variables are played against the left-hand side variable
* ebpcomp: new function to compute coordinates of components of an extended box plot. Along with spikecomp these are useful for adding things to ggplot2 graphs.
* spikecomp: new function to compute coordinates of components of a spike histogram
* movStats: new general function for estimating the relationship between a continuous variable and a response, possibly stratified by another variable
* csv.get: changed autodates to autodate and passed to cleanup.import; implemented text=
* fImport: new function; front-end for `rio` package for general file import
* transcan.Rd: added new example for completing a data frame with single imputations
* combine.levels: added plevels argument and implemented new capabilities for ordered variables, for which only consecutive levels are allowed to be combined
* fit.mult.impute: added fun argument so user can run any function on each fit from a completed dataset, e.g., a model validation using an rms function
* errbar: passed argument main to last instance of plot()
* princmp: fixed sparse PC total variance explained by asking for k PCs where k is one less than number of columns
* removed dependencies for Formula (importFrom), ggplot2 (importFrom), lattice (suggests), survival (suggests)
* NOTE: users who depended on ggplot2 or survival being loaded by Hmisc need to require() or library() themselves
* reformM: improved by Yong Hao Pua to handle variables enclosed in functions
* dotchartpl: fixed bug - was not sorting if groups omitted
* completer: new function by Yong-Hao Pua, Singapore General Hospital that facilitates drawing of multiple imputations to get one or more completed datasets
* ecdfSteps: compute coordinates of empirical CDF for a variable, with possible extension of the domain so that steps at y=0 and 1 will be visable
* fit.mult.impute: fixed to all data=a data table
(misc/R-Hmisc) Updated 4.7.0 to 4.8 0
Changes in version 4.8-0 (2023-02-08)
* rendHTML: new function to render HTML, sensing if knitr is
running and if so using knitr::asis_output, otherwise htmltools
* contents, describe: added the use of rendHTML so that HTML can be
rendered from the console, and made it so that if
options(prType='html') is in effect the html method will be
invoked automatically when printing
(misc/R-Hmisc) Updated 4.2.0 to 4.7.0, fix build against R-4.2.0
Changes in version 4.7-0 (2022-04-18)
* html.contents.data.frame: properly closed html <a>
* simPOcuts: new function to demonstrate variation in odds ratios
due to random chance
* R2Measures: new function to compute various pseudo R^2 measures
* putHcap: added new capabilities around the subsub argument
* print.summary.formula.response: added markdown argument
* knitrSet: added rudimentary quarto support
* knitrSet: sense figure labels of the form fig-... used by Quarto,
and generate correct cross-reference
Changes in version 4.6-0 (2021-10-05)
* package: improved author formatting in DESCRIPTION
* html: markupSpecs$html$session: added citations for any Harrell
packages that are loaded, respecting current output format in
effect with knitr
* soprobMarkovOrdm: new function to compute state occupancy
probabilities from proportion odds model fits
* plotCorrM: new function to graph correlation matrices and gap
time relationships using ggplot2
* ggplotlyr: new function to use plotly::ggplotly to render ggplot2
graphs but intercepting hover text to remove extraneous labels
* Fixed Heiberger email address
* propTrans: removed zero frequency combinations
* combplotp: fixed bug in case regarding recognition of positives
nnn * propsPO: fixed making y factor
* soprobMarkovOrdm: added as.data.frame(generated data)
* estSeqMarkovOrd: extended timecriterion function to allow user to
return the event/censoring time and indicator, and to allow
groupContrast to compute using a variance formula
* estSeqMarkovOrd: trapped errors better, returning attribute
failures, and added maxest and maxvest argument to declare large
parameter or variance estimates as failed iterations; changed
from vgam to vglm and sped up computations by using previous
coefficient estimates as starting values
* session in markupSpecs: added citations for several other packages
* soprobMarkovOrdm: extended to work with VGAM package
* describe: fix error with . (thanks to Cole Beck;
https://github.com/harrelfe/Hmisc/pull/144)
* many: tests for presence of suggested packages using
requireNamespace (thanks for major editing work by Duncan Murdoch
at https://github.com/harrelfe/Hmisc/pull/143 motivated by
https://stat.ethz.ch/pipermail/r-package-devel/2021q2/007101.html)
* mdb.get: fix for Windows by changing from system to system2
(thanks to Rainer Hurling at
https://github.com/harrelfe/Hmisc/pull/135)
* summaryM: fixed error with prN=TRUE with latex (thanks Matt
* Shotwell in https://github.com/harrelfe/Hmisc/pull/109)
* many: fix partial argument matching warnings length ->
* length.out in rep (thanks Bill Denney in
* https://github.com/harrelfe/Hmisc/pull/128)
* rcspline.restate: was dropping + for exactly zero coefficients
(thanks https://github.com/harrelfe/Hmisc/pull/118)
* mdb.get: remove brackets from table names (thanks
https://github.com/harrelfe/Hmisc/pull/123)
* latex: added new argument comment so that the generated comment
can be suppressed (thanks Giuseppe Ragusa
https://github.com/harrelfe/Hmisc/pull/33)
* sas.get: added new argument for variable case (thanks Tyler Hunt
https://github.com/harrelfe/Hmisc/pull/8)
* format.df: changed to use system option OutDec when cdot is not
specified (thanks https://github.com/harrelfe/Hmisc/issues/142)
* C code: changed calls to warning/error routines
* formatdescribeSingle: changed for character value to make 'NA' work
Changes in version 4.5-0 (2021-02-27)
* approxExtrap: changed x and y to as.numeric. Thanks: juha.heikkinen@luke.fi
* upData: fix column subsetting for data.tables
* dotchartpl: intercepted missing Diff
* simMarkovOrd, soprobMarkovOrd, intMarkovOrd, estSeqMarkovOrd: new
functions for Markov proportional odds model simulation and
calculation of state occupancy probabilities
* ggfreqScatter: added by argument
* VGAM package added to Suggests in DESCRIPTION
* html markupSpecs mdchunk: added caption argument, allowed for vectors
* propsTrans: added labels argument for plotly, added numerators
and denominators in tooltips, fixed bug where factor levels were
reversed if odds.ratio specified
Changes in version 4.4-2 (2020-11-25)
* rcorr: captured loss of precision leading to square root of a
negative number. Thanks: Ann Voss <avoss@novanthealth.org>
* summaryS: sapply was doubling names
* pairUpDiff: created for dotchartpl - function to pair up grouped
observations for sorting by descending differences and computing
approximate confidence intervals for the difference given
individual confidence limits
* dotchartpl: added new arguments and functionality to handle
quantities other than proportions (e.g., hazards) and streamlined
code using pairUpDiff
* propsPO: added tooltip text for gpplot that will be transmitted
to ggplotly; reversed order of categories so lowest category put
at bottom of bar
* dotchartpl: suppressed confidence intervals when the number of
groups is not 2; fixed bug where hover text confidence intervals
were repeats of the last computed interval instead of properly
using all the intervals; added dec argument
* added estSeqSim and gbayesSeqSim functions
* ggfreqScatter: stopped varying alpha and just varied color, as
alpha collided with the color scheme
* histSpike: added minimal=TRUE and bins arguments
Changes in version 4.4-1 (2020-08-07)
* popower: added approximate S.E. of log(OR) in results
* propsPO: new function for exploring proportional odds
* propsTrans: new function for showing distributions of successive
state transitions
* changed acepack to suggests
* multEventChart: new function for multi-state event charts based
on code written by Lucy D'Agostino McGowan
* getHdata: changed to use hbiostat.org/data/repo
* markupSpecs$markdown$tof: new function to render a table of
figures with short captions
* knitrSet: added capfile argument to store figure tags and short captions
* getLatestSource: changed to use GitHub and hbiostat.org/R/packagename/dir.txt
* histboxpM: added width argument
* upData: handled zero-length subsets (Thanks: Will Gray)
* upData: made faster for large datasets
* ffCompress: removed from package and put in Github Rscripts
* Changed maintainer email
Changes in version 4.4-0 (2020-03-22)
* combplotp: new function for attribute plots with plotly
* summaryP: made to work with new default stringsAsFactors=FALSE
* plotlyM: fixed bug where need to unlist if only one graph produced
* plotlyM: added ECDF support through fitter='ecdf'
* keepHattrib, restoreHattrib: new functions for saving Hmisc
attributes that can be restored later, e.g., after data.table
processes a data frame
Changes in version 4.3-1 (2020-02-07)
* Depend on survival >= 3.1-6
* pomodm: checked that x is sorted
* markupSpecs: added ord function for ordinal representation of integers
* mChoice: removed unused argument sort. from help file
Changes in version 4.3-0 (2019-11-07)
* Corrected typos in aregImpute help file. Thanks: Mark Seeto.
* describe: changed to print 5 lowest and highest values even if
frequency table printed, added a line of printed output
specifying any rounding done for the frequency table
* vcov.fit.mult.impute: use vcov.orm if object has orm class.
intercepts=mid logic was not working for this
* New service function convertPdate to do automatic date
conversions and handle partial dates such as YYYY and mm/YYYY
with imputation, for cleanup.import
* cleanup.import: new arguments autodate, autonum, fracnn
* describe: formatted dates if there is only one distinct value;
changed to sense date or date-times, for Gmd not format as
date-time
* plot.describe: treated date/time variables as numeric
* cleanup.import: refined autonum considerNA
* ggfreqScatter: added stick=TRUE argument
* markupSpecs: changed math and similar functions to use
... argument instead of x
* summaryDp: new function for plotly dotcharts stratifying
separately on a series of variables
* dotchartpl: added height argument
* html.describe: fixed but where markupSpecs$html utility function
last arguments were not named in calls
misc: Replace RMD160 checksums with BLAKE2s checksums All checksums have been double-checked against existing RMD160 and SHA512 hashes The following distfiles could not be fetched (possibly fetched conditionally?): ./misc/libreoffice/distinfo libreoffice/harfbuzz-2.6.4.tar.xz
misc: Remove SHA1 hashes for distfiles
R-Hmisc: initial commit. Contains many functions useful for data analysis, high-level graphics, utility operations, functions for computing sample size and power, importing and annotating datasets, imputing missing values, advanced table making, variable clustering, character string manipulation, conversion of R objects to LaTeX and html code, and recoding variables.