des.print: Create or print an output of a descriptive statistics table

In imbi-heidelberg/DescrTab: Creates a table with descriptive statistics

Description

This is the main function for the user. It can be used to generate a table of descriptive statistics with p-values obtained in tests for difference between the groups. There are five options for the output: pdf, tex, knitr, word file or an Output in R. knitr is not really useful as an output file but for the use within a knitr file.

Usage

des.print(
  dat,
  group,
  create = "knitr",
  file,
  index = T,
  fsize = 11,
  paired = F,
  nonparametric = F,
  var.equal = T,
  correct.cat = F,
  correct.wilcox = T,
  t.log = c(),
  which.col = c("groups", "total", "p-values"),
  group.min.size = F,
  group.non.empty = F,
  cat.non.empty = F,
  n.or.miss = c("n", "miss", "miss.cat"),
  adaptive.miss = T,
  group.miss = F,
  percent.vertical = T,
  var.names,
  data.names = T,
  caption,
  tab.caption,
  landscape = F,
  pos.pagebr = NULL,
  label = NULL,
  digits.m = 1,
  digits.sd = 2,
  digits.qu = c(),
  digits.minmax = 1,
  digits.p = c(1),
  silent = T,
  q.type = 2,
  alignment = "c",
  default.unordered.unpaired.test = "Chisq",
  ...
)

Arguments

dat	Data frame. The data set to be analyzed. Can contain continuous or factor (also ordered) variables.
group	The (optional) grouping variable. Three options to specify it: 1. The index of the group variable in dat. 2. The variable name. It must be a variable in dat. 3. As a vector with the same length as the number of rows in dat. 4. Nothing. Then you need which.col = ("total"). The specified variable has to be a factor variable with two or more levels. If not specified, a random grouping variable with 2 groups is used.
create	Which output document should be produced (one of "pdf", "tex", "knitr","word" or "R"). Choose "custom" if you add more arguments see ....
file	File name, which can included the directory ( has to have the proper file extension, i.e. .pdf, .tex, or .docx). directory. Only for create == "R" or "knitr" isn't a file necessary.
index	Logical. Should the tests used be labeled by footnotes?
fsize	The fontsize in the Output-Document. Default fontsize: For create = "word" or "R" fsize = 11 and for create = "pdf" fsize = 12
paired	Logical. Should paired tests be applied? The groups must have the same length.
nonparametric	Logical or vector of indices. If logical / vector of indices then all / only these continuous variables will be tested using non-parametric methods.
var.equal	Logical. Should variances be assumed to be equal when applying t-tests?
correct.cat	Logical. Should correction be used in chi-sqared tests (see chisq.test)
correct.wilcox	Logical. Should correction be used in wilcoxon tests (see wilcox.test)
t.log	Vector of indices: The variables for which the log of the original data should be used when testing for a difference between the groups.
which.col	Which columns should be provided ("groups", "total", "p-values")? Combinations are allowed. "groups" or "total" must be listed. Only "total" and "p-values" is not possible. Type of p-value calculation see p.cat (categorical) or p.cont (continuous).
group.min.size	For each variable, a p-value is only calculated if each non-empty group contains at least group.min.size observations for that variable.
group.non.empty	For each variable, a p-value is only calculated if each group contains at least one observation for that variable.
cat.non.empty	For categorical variables a p-value is only calculated if each category is non-empty.
n.or.miss	Should the number of observations, missings for continuous variables, and/or missings for categorical variables be provided ("n", "miss", "miss.cat")? Combinations are allowed.
adaptive.miss	Should the missing row be automatically omitted if there are not missings?
group.miss	Logical. Should add a column for the Missings in group? For both cases in total there are only the observations without a missing in group.
percent.vertical	Logical. Should "vertical" percentages for categorical variables be provided?
var.names	Optional. Vector of names to be used in the table for the analyzed variables.
data.names	Logical. If var.names are specified, should names as saved in dat be added in brackets?
caption	Optional. Vector of names of the categories of group. If not specified, levels(group) will be used.
tab.caption	Optional. The caption of the table. If not specified, "Descriptive statistics" will be written.
landscape	Logical. Should the table be in landscape? Only useful for create = "pdf".
pos.pagebr	Vector of positions of the pagebreak in tex (or pdf). This is a bit fuzzy. It is the number of lines after a pagebreak should be done. If it is not specified, 45 will be used for "landscape = FALSE" and 30 will be used for "landscape = TRUE".
label	Character. If tex (or knitr) in create is specified, the label of the table.
digits.m	Number of digits for presentation in the table: For mean.
digits.sd	Number of digits for presentation in the table: For standard deviation.
digits.qu	Vector of numbers of digits for presentation in the table: For quantiles (if no value is specified it will be tried to provide a reasonable presentation).
digits.minmax	Number of digits for presentation in the table: For minimum and maximum.
digits.p	Vector with numbers of digits for presentation in the table: For percentages. First vector element is number of digits for the first variable, second element for second variable and so on.
silent	Logical. Should intermediate stages be shown (more for technical reasons)?
q.type	an integer between 1 and 9 selecting one of the nine quantile algorithms detailed in ?quantile to be used.
default.unordered.unpaired.test	Any of c("Chisq", "Fisher_exact", "Fisher_boschloo"). Chooses the default test for categorical, unordered, unpaired variables. @param alignment Specifies where the table shoould be aligned in the ouput document. Default is "c", other options are "l" and "r"
...	further arguments to be passed to or from methods.

Details

The aim of this function is to help the user to create well-formated descriptive statistics tables. The format can then be fine-tuned in the word, tex or knitr file itself. Only the pdf file is (more or less) unchangeable. Even though it is supposed to be a good starting point for a descriptive analysis, are has to be taken when using the results and a detailed check of the results might be necessary. or instance, in case of missing values in the group variable and group.miss = FALSE the respective observations will be omitted also in the total column. A warning will be displayed. If no group variable is specified only the total column (see parameter which.col) will be returned.

Value

Only by using create == "R" an R object will be returned. However, depending on the value of the create parameter either pdf, word, tex or an file optimized for use in connection with knitr will be created containing the descriptive statistics table and saved in a file as specified in the file parameter.

Author(s)

Lorenz Uhlmann, Csilla van Lunteren

See Also

descr
f.r
formatr
inqur
m.cat
m.cont
med.new
minmax
p.cat
p.cont
autofit
flextable
set_header_df
merge_h
merge_at
align
style
border
bold
width
height
body_add_flextable
fp_text
fp_cell
fp_border
read_docx
xtable
texi2dvi

Examples

## Not run: 
infert
attach(infert)
#is.factor(education)
#is.factor(age)
#is.factor(parity)
#is.factor(induced)
#is.factor(case)
#infert$case<-as.factor(case)
#case<-as.factor(case)

#is.factor(spontaneous)
#is.factor(stratum)
#is.factor(pooled.stratum)

#we use case as Grouping variable
file <- "C:/Users/lunteren/Desktop/infert/DescriptiveStatisticTable.docx"

#Version 1
des.print(dat = infert, group = 5, create = "word", file=file, fsize = 10,
var.names = c("education", "age", "parity","induced", "spontaneous", "stratum", "pooled.stratum"),
caption = c("Group 1", "Group 2"))
#Version 2
des.print(dat = infert, group = case, create = "word", file=file, fsize = 10,
var.names = c("education", "age", "parity","induced", "spontaneous", "stratum", "pooled.stratum"),
caption = c("Group 1", "Group 2"))
#Version 3
group <- case
dat <- infert[,-5]
des.print(dat = dat, group = group, create = "word", file=file, fsize = 10,
var.names = c("education", "age", "parity","induced", "spontaneous", "stratum", "pooled.stratum"),
caption = c("Group 1", "Group 2"))

##Dataset with more then two groups
ChickWeight
attach(ChickWeight)

#is.factor(weight)
#is.factor(Time)
#is.factor(Chick)
#is.factor(Diet)

#we use Diet as Grouping variable

file <- "C:/Users/lunteren/Desktop/ChickWeight/DescriptiveStatisticTable.docx"

#Version 1
des.print(dat = ChickWeight, group = 4, create = "word", file = file, fsize = 10,
var.names = c("weight", "Time", "Chick"),
caption = c("Group 1", "Group 2", "Group 3", "Group 4"))
#Version 2
des.print(dat = ChickWeight, group = Diet, create = "word", file = file, fsize = 10,
var.names = c("weight", "Time", "Chick"),
caption = c("Group 1", "Group 2", "Group 3", "Group 4"))
#Version 3
group <- Diet
dat <- ChickWeight[,-4]
des.print(dat = dat, group = group, create = "word", file = file, fsize = 10,
var.names = c("weight", "Time", "Chick"),
caption = c("Group 1", "Group 2", "Group 3", "Group 4"))

## End(Not run)

imbi-heidelberg/DescrTab documentation built on June 29, 2021, 2:07 p.m.