LinearizeNestedList: Linearize (Un-nest) Nested Lists
In mrdwab/mrdwabmisc: Miscellaneous R functions, mostly for data processing

Description Usage Arguments Details Author(s) References See Also Examples

Implements a recursive algorithm to linearize nested lists upto any arbitrary level of nesting (limited by R's allowance for recursion-depth). By linearization, it is meant to bring all list branches emanating from any nth-nested trunk upto the top-level trunk such that the return value is a simple non-nested list having all branches emanating from this top-level branch.

1 2	LinearizeNestedList(NList, LinearizeDataFrames = FALSE, NameSep = "/", ForceNames = FALSE)

`NList`	The input `list`
`LinearizeDataFrames`	Logical. Should columns in `data.frame`s in the list be "linearized" as vectors? Defaults to `FALSE`.
`NameSep`	Character to be used when creating names. Defaults to "/" to mimic directory listings.
`ForceNames`	Logical. Should the present names be discarded and new simplified names be created? Defaults to `FALSE`

Since data.frames are essentially lists a boolean option is provided to switch on/off the linearization of data.frames. This has been found desirable in the author's experience.

Also, one would typically want to preserve names in the lists in a way as to clearly denote the association of any list element to its nth-level history. As such we provide a clean and simple method of preserving names information of list elements. The names at any level of nesting are appended to the names of all preceding trunks using the NameSep option string as the seperator. The default "/" has been chosen to mimic the unix tradition of filesystem hierarchies. The default behavior works with existing names at any n-th level trunk, if found; otherwise, coerces simple numeric names corresponding to the position of a list element on the nth-trunk. Note, however, that this naming pattern does not ensure unique names for all elements in the resulting list. If the nested lists had non-unique names in a trunk the same would be reflected in the final list. Also, note that the function does not at all handle cases where some names are missing and some are not.

Clearly, preserving the n-level hierarchy of branches in the element names may lead to names that are too long. Often, only the depth of a list element may only be important. To deal with this possibility a boolean option called ForceNames has been provided. ForceNames shall drop all original names in the lists and coerce simple numeric names which simply indicate the position of an element at the nth-level trunk as well as all preceding trunk numbers.

Akhil S Bhel

https://sites.google.com/site/akhilsbehl/geekspace/articles/r/linearize_nested_lists_in_r

unlist

# Create some sample data
NList <- list(a = "a", # Atom
     b = 1:5, # Vector
     c = data.frame(x = runif(5), y = runif(5)), # Add a data.frame
     d = matrix(runif(4), nrow = 2), # Throw in a matrix for good measure
     e = list(l = list("a", "b"), # Introduce nesting
              m = list(1:5, 5:10),
              n = list(list(1), list(2)))) # More nesting

LinearizeNestedList(NList)
LinearizeNestedList(NList, LinearizeDataFrames = TRUE)
LinearizeNestedList(NList, ForceNames = TRUE)
LinearizeNestedList(NList, ForceNames = TRUE, NameSep = "_")