make.table: Make Summary Table

In emwozniak/Table1: Summary Table

Make Summary Table

Description

This function produces summary statistics for categeorical and continuous variables and combines the results into a data frame for plain text, HTML, or LaTeX output, serving as a wrapper to apply vectorized inputs to cat.var and cont.var. The input dataset must be cleaned (e.g. blank cells and NaNs replaced with NAs).

Usage

make.table(dat, cat.varlist, cat.header, cat.rownames, cat.ptype,
cont.varlist, cont.header, cont.ptype, strat, cat.rmstat, cont.rmstat,
dec, pname, colnames, output, vspace, varorder, stripe, stripe.col, header.style,
factor.style, stat.style, nowrap, caption, footer, tspanner, n.tspanner,
cgroup, n.cgroup, col.columns)

Arguments

dat	Clean data frame input of all variables used for the summary table.
cat.varlist	Vector of categorical variable names. Optional; defaults to NULL.
cat.header	Ordered vector of short descriptions for each categorical variable. Optional; defaults to variable names.
cat.rownames	Ordered list of vectors providing row names for each level of each vector found in cat.varlist. Optional; defaults to the coded value of each level.
cat.ptype	Ordered vector of tests requested to add p-values to table for categorical variables. Optional; defaults to no p-values. See stat.col for a complete list of options.
cont.varlist	Vector of continuous variable names. Optional; defaults to NULL.
cont.header	Ordered vector of short descriptions for each continuous variable. Optional; defaults to variable name.
cont.ptype	Ordered vector of tests requested to add p-values to table for continuous variables. Optional; defaults to no p-values. See stat.col for a complete list of options.
strat	Vector of one or more stratifying variables. Optional; defaults to no stratification.
cat.rmstat	Ordered list of vectors of statistics that should be removed from respective final summaries of categorical variables.
cont.rmstat	Ordered list of vectors of statistics that should be removed from respective final summaries of continuous variables.
dec	Integer value indicating the number of decimal places to be printed for each summary statistic. Optional; defaults to 2.
pname	Logical indicator for whether to print statistical test used beneath each p-value. Optional; defaults to TRUE.
colnames	Ordered vector of column names for the table. Optional; defaults to levels of stratifying variable(s) with an overall summary column.
output	Choice of formatted table output: "html", "latex", or "plain". Optional; defaults to plain text printed to the console.
vspace	TRUE for additional vertical whitespace between variables (default). Optional.
varorder	Order of variables in the outputted table. Defaults to "data", or the order of variables in the inputted dataset. Alternatively, variables may be presented alphabetically using the "abc" option. Optional.
stripe	HTML output only: logical indicator for whether to zebra stripe every other variable in the table. Optional; defaults to TRUE.
stripe.col	HTML output only: hex color code to be used for zebra striping the table. Optional; defaults to a light gray, #F7F7F7.
header.style	Variable name style: "plain", "bold", "italic", or "bolditalic". Optional; defaults to "bold".
factor.style	Variable factor levels style: "plain", "bold", "italic", or "bolditalic". Optional; defaults to "bold".
stat.style	Summary statistic style: "plain", "bold", "italic", or "bolditalic". Optional; defaults to "plain".
nowrap	HTML output only: option to prevent text wrapping of long lines to the table output. Useful for generating wide tables with knitr. Optional; defaults to TRUE.
caption	Table caption. Optional; no default.
footer	Table footer. Optional; no default.
tspanner	HTML output only: Rowwise table spanner, passed to htmlTable. Optional; see ?htmlTable for default information.
n.tspanner	HTML output only: Vector of rows in each table spanner, passed to htmlTable. Optional; see ?htmlTable for default information.
cgroup	HTML output only: Column group headers, passed to htmlTable. Optional; see ?htmlTable for default information.
n.cgroup	HTML output only: Vector of columns under each group header, passed to htmlTable. Optional; see ?htmlTable for default information.
col.columns	HTML output only: Vector of column colors, passed to htmlTable. Optional; defaults to "none".

Value

Returns a table of summary statistics for specified categorical and continuous variables in the requested output format. The plain text default returns a formatted data frame to the console.

Note

A few miscellaneous notes:

Table output to HTML and LaTeX require the package dependencies htmlTable and xtable

If missing values occur in stratifying variables (strat =), then the observations with missings will be removed from the table, and a comment will be printed indicating the total number removed. If multiple stratifying variables are provided, look to the console for a summary of total missing observations for each variable.

Author(s)

Erica Wozniak

See Also

Table1 quick.table cat.var cont.var stat.col out.plain out.latex out.html

Examples

library(survival)

# Example with non-HTML options
make.table(dat   = pbc,
    cat.varlist  = c('stage', 'sex'),
    cat.header   = c('Stage', 'Sex'),
    cat.rownames = list(c('I', 'II', 'III', 'IV'), c('Male', 'Female')),
    cat.ptype    = c('fisher', 'fisher'),   
    cont.varlist = c('bili', 'copper'),
    cont.header  = c('Bilirubin', 'Urine copper'),
    cont.ptype   = c('wilcox', 'wilcox'), 
    strat        = c('trt'),
    # Remove row percents from all categorical variables (list is recycled)
    cat.rmstat   = list('row'),
    # Remove missing, min/max from bili and no summary stats from copper
    cont.rmstat  = list(c('miss', 'minmax'), 'None'),
    colnames     = c('', 'D-penicillamine', 'Placebo', 'Overall', 'p-value')
    )
    
# Example with HTML options
library(htmlTable)
make.table(dat   = pbc,
    cat.varlist  = c('stage', 'sex'),
    cat.header   = c('Stage', 'Sex'),
    cat.rownames = list(c('I', 'II', 'III', 'IV'), c('Male', 'Female')),
    cat.ptype    = c('fisher', 'fisher'),   
    cont.varlist = c('bili', 'copper'),
    cont.header  = c('Bilirubin', 'Urine copper'),
    cont.ptype   = c('wilcox', 'wilcox'), 
    strat        = c('trt'),
    cat.rmstat   = list('row'),
    colnames     = c('', 'D-penicillamine', 'Placebo', 'Overall', 'p-value'),
    output       = 'html',   
    # Check out \url{www.color-hex.com}
    stripe.col = '#f4dfd0'
    )

emwozniak/Table1 documentation built on Oct. 5, 2022, 12:20 a.m.