equilibrium-causal-models

https://github.com/fdabl/equilibrium-causal-models

Repository

README.html

Equilibrium Causal Models
This repository contains code to reproduce all analyses and figures
in Ryan, O.⭑ & Dablander, F.⭑ (2022). Equilibrium Causal Models:
Connecting Dynamical Systems Modeling and Cross-Sectional Data
Analysis.

figures.R code to reproduce all figures in the
manuscript
ECM.R helper functions relating to ECMs, such as
computing shift and press interventions, re-scaling matrices, etc.
helpers.R useful R functions for simulation and
plotting
Simulation/ holds R code to reproduce all
simulation studies in the paper:

Simulation/simulation_sanity.R runs the simulation
summarized in Figure 4.
Simulation/simulation_measurement.R runs the simulation
summarized in Figure 6.
Simulation/simulation_backshift_create.R creates data
and estimates causal models using backshift.
Simulation/simulation_backshift_analyze.R calculates
metrics of estimated causal models (summarized in Figures 10 and
11)

Results/ includes all simulation results discussed
in the manuscript:

Results/sanity_results.RDS holds the results of the
simulation summarized in Figure 4.
Results/measurement_results.RDS holds the results of
the simulation summarized in Figure 6.
Results/state_trait_results.RDS holds the results of
the simulation summarized in Figure 7.
Results/backshift_estimates.RDS holds the estimated
causal models.
Results/backshift_metrics.csv holds the metrics
calculated on the estimated causal models.

Example/ includes code for simulated and empirical
examples using lavaan and backShift:

Example/example_datagen.R creates simulated data based
on the running example in the main text with imperfect measurements of
the equilibrium, used by example_modelfit.R.
Example/example_data.RDS simulated data set created by
example_datagen.R.
Example/example_modelfit.R fits the ECM using
lavaan and the measurement model constraints described in
Appendix C.
Example/example_backshift.R creates equilibrium data
and estimates an ECM using backShift, as in the simulation
described in Appendix D.
/EmpiricalExample/ includes the code for the
empirical examples in appendices D and F

empirical_example_mcneish.R illustrates the estimation
of an ECM from multilevel time series and estimated equilibriums
respectively using data from McNeish and MacKinnon (2022) as described
in Appnedix D. It loads Mplus model output objects
stationary mediation lag1.out, along with
bpars_saved_l1.dat and fscores_saved_l1.dat.
Data is open access from https://osf.io/yk3je/.
empirical_example_backshift.R illustrates using
backShift on the data of Blanken et al. (2019), as
described in Appendix F.


Figures/ includes all figures in the
manuscript.

Owner

Citation (CITATION.cff)

# This CITATION.cff file was generated with cffinit.
# Visit https://bit.ly/cffinit to generate yours today!

cff-version: 1.2.0
message: "Supplementary Material for Equilibrium Causal Models: Connecting Dynamical Systems Modeling and Cross-Sectional Data Analysis"
authors:
  - given-names: Fabian
    family-names: Dablander
    email: f.dablander@uva.nl
    affiliation: University of Amsterdam
  - given-names: Oisín
    family-names: Ryan
    email: o.ryan@uu.nl
    affiliation: Utrecht University
title: "Supplementary Material for Equilibrium Causal Models: Connecting Dynamical Systems Modeling and Cross-Sectional Data Analysis"
version: 1.0.0
date-released: 2022-11-02
url: "https://github.com/fdabl/Equilibrium-Causal-Models"

GitHub Events