Two simulated datasets for demonstrating growth models on SSM parameters with
ssm_parameters_id() and ssm_draws(). The data are simulated, not collected
from people. In both datasets, 150 persons have scores on the eight octant
scales at waves 0 to 4.
Format
Each is a data frame with 750 rows (one per person per wave) and 10 columns:
- person
Person identifier, 1 to 150.
- wave
Wave number, an integer from 0 to 4.
- PA, BC, DE, FG, HI, JK, LM, NO
Scores on the eight octant scales, at the angles given by
octants().
An object of class data.frame with 750 rows and 10 columns.
Details
A person's score on a scale at a wave is the elevation 0.5, plus a person elevation effect, plus \((x + v_x)\cos\theta + (y + v_y)\sin\theta\), plus noise. Here \(\theta\) is the scale angle and \((x, y)\) is the group coordinate at that wave. The person elevation effect has standard deviation 0.30, the person effects \(v_x\) and \(v_y\) have standard deviation 0.15, and the noise has standard deviation 0.40. All of them are drawn from normal distributions. The person effects stay the same at every wave, and the two datasets share them, so each person has the same effects in both.
The two datasets differ in the group coordinates:
In
simulated_growth, \((x, y)\) moves in a straight line, in equal steps, from the point with amplitude 0.6 at 350 degrees to the point with amplitude 0.6 at 10 degrees. The displacement crosses the 0/360 degree boundary.In
simulated_growth_origin, \(x\) moves in equal steps from 0.5 to -0.5, and \(y\) stays at 0.02. The group passes near the origin at wave 2.
Both datasets are written by one seeded script, data-raw/simulated_growth.R,
which is not in the installed package. Read it on GitHub:
https://github.com/jmgirard/circumplex/blob/master/data-raw/simulated_growth.R.