preview: Output variable states within functions during...

In NCmisc: Miscellaneous Functions for Creating Adaptive Functions and Scripts

Output variable states within functions during testing/debugging

Description

A versatile function to compactly display most common R objects. Will return the object name, type, dimension, and a compact representation of object contents, for instance using prv.large() to display matrices, so as to not overload the console for large objects. Useful for debugging, can be placed inside loops and functions to track values, dimensions, and data types. Particularly when debugging complex code, the automatic display of the variable name prevents confusion versus using regular print statements. By listing variables to track as character(), provides 'cat()' output of compact and informative variable state information, e.g, variable name, value, datatype and dimension. Can also specify array or list elements, or custom labels. prv() is the same as preview() except it can take objects without using double quotes and has no 'labels' command (and doesn't need one).

Usage

preview(
  varlist,
  labels = NULL,
  counts = NULL,
  assume.char = FALSE,
  prv.call = FALSE
)

Arguments

varlist	character vector, the list of variable(s) to report, which will trigger automatic labelling of the variable name, otherwise if entered as the variable value (ie. without quotes, then will by default be displayed as 'unknown variable')
labels,	will label 'unknown variables' (see above) if entered as variables without quotes
counts	a list of array index values; so if calling during a counting loop, the value can be reported each iteration, also printing the count index; if the list is named the name will also appear, e.g, variable[count=1]. This list must be the same length as varlist (and labels if not NULL), and each element [[i]] must contain as many values as the original corresponding varlist[i] has dimensions. The dimensions must result in a 1x1 scalar
assume.char	usually 'varlist' is a character vector of variable names, but in the case that it is actually a character variable, using assume.char=TRUE will ensure that it will be assumed the character variable is the object to preview, rather than a list of variable names. So long as none of the values are found to be variable names in the global environment. preview() can also find variables in local environments, and if this is where the target variable lies, it is best to use assume.char=FALSE, otherwise the search for alternative environments might not happen. Note that in most cases the automatic detection of the input should understand what you want, regardless of the value of assume.char.
prv.call	It is recommended to always leave this argument as FALSE when calling preview() directly. If set to TRUE, it will first search 2 generations back for the parent frame, instead of one, as it will assume that the variable(s) to preview are not directly called by preview(), but through a wrapper for preview, such as prv().

See Also

Dim

Examples

# create variables of different types to show output styles #
testvar1 <- 193
testvar2 <- "Atol"
testvar3 <- c(1:10)
testvar4 <- matrix(rnorm(100),nrow=25)
testvar5 <- list(first="test",second=testvar4,third=100:110)
preview("testvar1")
preview("testvar4")
preview(paste("testvar",1:5,sep=""))
preview(testvar1,"myvarname")
preview(testvar1)
# examples with loops and multiple dimensions / lists
for (cc in 1:4) {
 for (dd in 1:4) { preview("testvar4",counts=list(cc,dd)) }}

for (dd in 1:3) { preview("testvar5",counts=list(dd=dd)) }

NCmisc documentation built on Oct. 17, 2022, 5:09 p.m.