search

milanwiedemann / lcsm

An R package for latent change score modelling (LCSM)
https://milanwiedemann.github.io/lcsm/
Other
17 stars 3 forks source link

add function to automatically define layout matrix for semPlot #2

Closed milanwiedemann closed 4 years ago

milanwiedemann commented 5 years ago

at the moment the matrix needs to be specified manually, see the following for a univariate an bivariate lcsm with seven time points.

# Create layout matrix for univariate LCSM without intercepts
layout_uni_lcsm <- matrix(
  c(NA, "g2", NA, NA, NA, NA, NA, 
    NA, "dx2", "dx3", "dx4", "dx5", "dx6", "dx7",
    "lx1", "lx2", "lx3", "lx4", "lx5", "lx6", "lx7",
    "x1", "x2", "x3", "x4", "x5", "x6", "x7"),
  4, byrow = TRUE)

# Create layout matrix for bivariate LCSM without intercepts
layout_bi_lcsm <- matrix(
  c(NA, NA, "y1", "y2", "y3", "y4", "y5", "y6", "y7",
    NA, NA, "ly1", "ly2", "ly3", "ly4", "ly5", "ly6", "ly7",
    NA, NA, NA, "dy2", "dy3", "dy4", "dy5", "dy6", "dy7",
    NA, "j2",NA,NA, NA,  NA, NA, NA, NA, 
    NA, "g2", NA,NA, NA,  NA, NA, NA, NA, 
    NA, NA, NA, "dx2", "dx3", "dx4", "dx5", "dx6", "dx7", 
    NA, NA, "lx1", "lx2", "lx3", "lx4", "lx5", "lx6", "lx7",
    NA, NA, "x1", "x2", "x3", "x4", "x5", "x6", "x7"),
  8, byrow = TRUE)
milanwiedemann commented 5 years ago

working on it, idea to create matrix from lavaan syntax, using lavaanify to get all info about manifest and latent vars and change factors

milanwiedemann commented 4 years ago

done a long time ago