formatting: Format numerical results for export tables

In CALIBERdatamanage: Data Management Tools for CALIBER Datasets

Description

Functions to format numbers for publication, producing either plain text or LaTeX output. formatnum is the general function. formatp formats P values. formathr exponentiates and prints coefficients such as hazard ratios with confidence intervals, and displays significance levels with stars.

Usage

formatnum(numbers, dp = NA, sigfig = 3, scientific = TRUE,
    latex = (getOption("xtable.type") == "latex"),
    usephantom = getOption("CALIBERdatamanage_phantom"), names = NULL,
    decimal.mark = ".")
formatp(pvalues, minp = 0.0001,
    latex=(getOption('xtable.type')=='latex'), ...)
formathr(coef, se, df = Inf, dp = NA, pstar = TRUE,
    pvalueci = 0.05, template = '\\1 (\\2, \\3)', ...)
formatci(range, scientific = FALSE, template = '(\\1, \\2)', ...)

Arguments

numbers	numeric vector of numbers to format
dp	number of decimal places
sigfig	number of significant figures
scientific	TRUE for scientific mode, FALSE for standard output
latex	TRUE for LaTeX output (math mode), FALSE, NULL or logical(0) for plain text output
usephantom	TRUE to pad the LaTeX output with blank (phantom) characters so that the decimal points line up
names	names for the output vector
pvalues	numeric vector of P value values to be formatted
minp	minimum P value to display as an exact value
coef	coefficient (log hazard ratio or log odds ratio)
se	standard error of coefficient
df	degrees of freedom for t-distribution; Inf to use the Normal distribution
pstar	TRUE to include stars for P value: * < 0.05, ** < 0.01, *** < 0.001
range	numeric vector containing the upper and lower confidence limits
pvalueci	significance level to be used in the calculation of confidence intervals
template	character string template with \1 and \2 representing numbers forming the confidence interval (formatci), and \1, \2 and \3 representing the estimate and confidence interval for formathr.
decimal.mark	character to be used for decimal point, default is .
...	other arguments to be passed to formatnum

Value

a character vector containing formatted numbers

Author(s)

Anoop Shah

See Also

percent, npercent, percentConf, nmissing, meansd

Examples

formatnum(1:5, latex = FALSE)
# [1] "1" "2" "3" "4" "5"
formatnum(1:5, latex = TRUE)
# [1] "$1$" "$2$" "$3$" "$4$" "$5$"
formatnum(c(0.00001, 0.02, 5.6, 67, 700), latex = FALSE, dp = 1)
# [1] "0.0"   "0.0"   "5.6"   "67.0"  "700.0"
formatnum(c(0.00001, 0.02, 5.6, 67, 700), latex = TRUE, usephantom = TRUE)
# [1] "$1\times 10^{-5}$"                       
# [2] "\phantom{0}\phantom{0}$0.02$"           
# [3] "\phantom{0}\phantom{0}$5.6\phantom{0}$"
# [4] "\phantom{0}$67\phantom{.}\phantom{0}\phantom{0}$"
# [5] "$700\phantom{.}\phantom{0}\phantom{0}$"

# Formatting hazard ratios
formathr(0.8, 0.3, latex = FALSE)
# "2.23 (1.24, 4.01) **"
formathr(0.8, 0.3, latex = TRUE)
# "$2.23$ ($1.24$, $4.01$) **"
formathr(0.8, 0.3, latex = TRUE, template = '\1; \1--\2')
# "$2.23$; $1.24$--$4.01$ **"
# Missing values should not cause errors
formathr(c(0.8, 0.8, NA), c(NA, 0.3, 0.3))
# [1] "2.23 (NA, NA)"        "2.23 (1.24, 4.01) **" "NA (NA, NA)" 

# Formatting confidence intervals
formatci(c(2.4343, 4.3245), latex = FALSE)
# "(2.43, 4.32)"
formatci(c(2.4343, 4.3245), latex = TRUE)
# "($2.43$, $4.32$)"

CALIBERdatamanage documentation built on Nov. 23, 2021, 3 p.m.