scrapeVarCross: A Variable Name-Scraping and Indexing Function for...

In jsakaluk/dySEM: Dyadic Structural Equation Modeling

A Variable Name-Scraping and Indexing Function for cross-sectional data

Description

This function scrapes the names of indicator variables in a wide-format data set used for dyadic analyses of two latent variables (LV; X and Y), and indexes which indicators correspond to which partner, for which LV. It is used primarily to guide the syntax-writing of the other dySEM functions.

Usage

scrapeVarCross(
  dat,
  x_order = "spi",
  x_stem,
  x_delim1 = NULL,
  x_delim2 = NULL,
  x_item_num = "\\d+",
  distinguish_1 = "1",
  distinguish_2 = "2",
  y_order = NULL,
  y_stem = NULL,
  y_delim1 = NULL,
  y_delim2 = NULL,
  y_item_num = "\\d+",
  var_list = NULL,
  var_list_order = NULL,
  var_list_item_num = "\\d+",
  covs_order = NULL,
  covs_stem = NULL,
  covs_delim1 = NULL,
  covs_delim2 = NULL
)

Arguments

dat	input data frame of indicators of a particular LV
x_order	input character for order of (S)tem, (P)artner number, and (I)tem number when creating variable names. Defaults to "spi" (Qualtrics-friendly), but can alternatively take "sip" or "psi"
x_stem	input character stem of indicator variables for LV X
x_delim1	optional character to separate stem from partner number (spi) or item number (sip)
x_delim2	optional character to separate stem/partner number (spi) or stem/item number (sip) from from final element of variable name
x_item_num	defaults to scrape all items that match the stem with any digits that follow. Will be updated to allow particular range of values, to make more sub-scale friendly.
distinguish_1	input character used as the identifier for the first partner
distinguish_2	input character used as the identifier for the first partner
y_order	optional character for order of (S)tem, (P)artner number, and (I)tem number when creating variable names. Defaults to "spi" (Qualtrics-friendly), but can alternatively take "sip" or "psi". This and other Y-arguments only necessary if there is a latent Y variable to model
y_stem	optional input character stem of indicator variables for LV X
y_delim1	optional character to separate stem from partner number (spi) or item number (sip)
y_delim2	optional character to separate stem/partner number (spi) or stem/item number (sip) from from final element of variable name
y_item_num	defaults to scrape all items that match the stem with any digits that follow. Will be updated to allow particular range of values, to make more sub-scale friendly.
var_list	optional named list of indicator variable information, if more than one LV is to be scripted (e.g., a dyadic CFA with multiple sub-scales from the same measure). If supplied, this list must contain the following elements: "stem" (a vector of stems), "delim1" (a vector of delimiting characters), and"delim2" (a vector of subsequently delimiting characters). Optionally may include numeric vectors "min_num" and "max_num" if indicators for different LVs share the same stem and must be separated by range of item numbers within a measure.
var_list_order	optional character for order of (S)tem, (P)artner number, and (I)tem number for any of the indicator variables of a multi-LV model (i.e., this functionality assumes the same ordering of elements throughout)
var_list_item_num	optional character for item number of any of the indicator variables of a multi-LV model
covs_order	optional character for order of (S)tem, (P)artner number, and (I)tem number for any covariate(s). Defaults to NULL. This and other covariate arguments only necessary if there are covariates to be scripted in your model(s).
covs_stem	optional input character stem(s) of indicator variables for covariate(s). Can accept a single stem (e.g., "anx"), or a vector of stems (e.g., c("anx", "dep")). Defaults to NULL.
covs_delim1	optional character to separate stem from partner number (spi) or item number (sip) for covariate(s). Defaults to NULL.
covs_delim2	optional character to separate stem/partner number (spi) or stem/item number (sip) from

Value

a list, referred in short-hand as a "dvn" (dyad variable names list) containing variable names for p1, p2, # of items per LV, characters distinguishing partners, and total number of indicators

Examples

dvnx <- scrapeVarCross(dat = commitmentQ, x_order = "spi", x_stem = "sat.g", x_delim1 = ".",
x_delim2="_", distinguish_1="1", distinguish_2="2")
dvnxy <- scrapeVarCross(dat = commitmentQ, x_order = "spi", x_stem = "sat.g", x_delim1 = ".",
x_delim2="_", distinguish_1="1", distinguish_2="2",
y_order="spi", y_stem="com", y_delim1 = ".", y_delim2="_")

jsakaluk/dySEM documentation built on Feb. 24, 2025, 5:05 p.m.