Fit the Discrete-Time Vector Autoregressive Model By ID (Adaptive Recovery

Dynamics Description

The Adaptive Recovery process reflects an asymmetric regulatory dynamic between two latent constructs—such as stress and coping—where activation in one system initiates a corrective response in the other. Specifically, stress tends to increase coping responses, while coping reduces subsequent stress, producing a negative feedback loop that promotes stability and recovery.

Individuals differ in the strength and balance of these cross-regulatory influences, leading to variability in how quickly they return to equilibrium after disturbances. The process noise covariance is moderate and negatively correlated, representing compensatory fluctuations where increases in stress are often accompanied by decreases in coping, while measurement error variance is small and symmetric across variables.

This configuration captures a psychologically meaningful stress-response mechanism, characterized by self-correcting dynamics that stabilize the system over time through coordinated but asymmetric influences.

Model

The measurement model is given by $\begin{equation} \mathbf{y}_{i, t} = \boldsymbol{\nu}_{i} + \boldsymbol{\Lambda} \boldsymbol{\eta}_{i, t} + \boldsymbol{\varepsilon}_{i, t}, \quad \mathrm{with} \quad \boldsymbol{\varepsilon}_{i, t} \sim \mathcal{N} \left( \mathbf{0}, \boldsymbol{\Theta} \right) \end{equation}$ where $\mathbf{y}_{i, t}$ , $\boldsymbol{\eta}_{i, t}$ , and $\boldsymbol{\varepsilon}_{i, t}$ are random variables and $\boldsymbol{\nu}_{i}$ , $\boldsymbol{\Lambda}$ , and $\boldsymbol{\Theta}$ are model parameters. $\mathbf{y}_{i, t}$ represents a vector of observed random variables, $\boldsymbol{\eta}_{i, t}$ a vector of latent random variables, and $\boldsymbol{\varepsilon}_{i, t}$ a vector of random measurement errors, at time $t$ and individual $i$ . $\boldsymbol{\Lambda}$ denotes a matrix of factor loadings, and $\boldsymbol{\Theta}$ the covariance matrix of $\boldsymbol{\varepsilon}$ that is invariant across individuals. In this model, $\boldsymbol{\Lambda}$ is an identity matrix and $\boldsymbol{\Theta}$ is a symmetric matrix.

The dynamic structure is given by $\begin{equation} \boldsymbol{\eta}_{i, t} = \boldsymbol{\beta}_{i} \boldsymbol{\eta}_{i, t - 1} + \boldsymbol{\zeta}_{i, t}, \quad \mathrm{with} \quad \boldsymbol{\zeta}_{i, t} \sim \mathcal{N} \left( \mathbf{0}, \boldsymbol{\Psi} \right) \end{equation}$ where $\boldsymbol{\eta}_{i, t}$ , $\boldsymbol{\eta}_{i, t - 1}$ , and $\boldsymbol{\zeta}_{i, t}$ are random variables, and $\boldsymbol{\beta}_{i}$ , and $\boldsymbol{\Psi}$ are model parameters. Here, $\boldsymbol{\eta}_{i, t}$ is a vector of latent variables at time $t$ and individual $i$ , $\boldsymbol{\eta}_{i, t - 1}$ represents a vector of latent variables at time $t - 1$ and individual $i$ , and $\boldsymbol{\zeta}_{i, t}$ represents a vector of dynamic noise at time $t$ and individual $i$ . $\boldsymbol{\beta}_{i}$ is a matrix of autoregression and cross regression coefficients for individual $i$ , and $\boldsymbol{\Psi}$ the covariance matrix of $\boldsymbol{\zeta}_{i, t}$ that is invariant across all individuals. In this model, $\boldsymbol{\Psi}$ is a symmetric matrix.

Data Generation

Notation

Let $t = 100$ be the number of time points and $n = 100$ be the number of individuals. We simulate a total of time $= 10100$ points per individual, discarding the first $10000$ as burn-in. The analysis uses the final $100$ measurement occasions.

Let the measurement model intercept vector $\boldsymbol{\nu}$ be normally distributed with the following means

$\begin{equation} \left( \begin{array}{c} 0.8 \\ 0.5 \\ \end{array} \right) \end{equation}$

and covariance matrix

$\begin{equation} \left( \begin{array}{cc} 0.15 & 0.05 \\ 0.05 & 0.1 \\ \end{array} \right) . \end{equation}$

Let the factor loadings matrix $\boldsymbol{\Lambda}$ be given by

$\begin{equation} \boldsymbol{\Lambda} = \left( \begin{array}{cc} 1 & 0 \\ 0 & 1 \\ \end{array} \right) . \end{equation}$

Let the measurement error covariance matrix $\boldsymbol{\Theta}$ be given by

$\begin{equation} \boldsymbol{\Theta} = \left( \begin{array}{cc} 0.5 & 0 \\ 0 & 0.5 \\ \end{array} \right) . \end{equation}$

Let the initial condition $\boldsymbol{\eta}_{0}$ be given by

$\begin{equation} \boldsymbol{\eta}_{0} \sim \mathcal{N} \left( \boldsymbol{\mu}_{\boldsymbol{\eta} \mid 0}, \boldsymbol{\Sigma}_{\boldsymbol{\eta} \mid 0} \right) . \end{equation}$

$\boldsymbol{\mu}_{\boldsymbol{\eta} \mid 0}$ and $\boldsymbol{\Sigma}_{\boldsymbol{\eta} \mid 0}$ are functions of $\boldsymbol{\alpha}$ and $\boldsymbol{\beta}$ .

Let the transition matrix $\boldsymbol{\beta}$ be normally distributed with the following means

$\begin{equation} \left( \begin{array}{cc} 0.6 & -0.3 \\ 0.25 & 0.7 \\ \end{array} \right) \end{equation}$

and covariance matrix

$\begin{equation} \left( \begin{array}{cccc} 0.025 & 0.01 & 0 & 0 \\ 0.01 & 0.02 & 0 & 0 \\ 0 & 0 & 0.02 & 0.01 \\ 0 & 0 & 0.01 & 0.025 \\ \end{array} \right) . \end{equation}$

Let the intercept vector $\boldsymbol{\alpha}$ be fixed to a zero vector.

The SimNuN and SimBetaN functions from the simStateSpace package generate random intercept vectors and transition matrices from the multivariate normal distribution. Note that the SimBetaN function generates transition matrices that are weakly stationary with an option to set lower and upper bounds.

Let the dynamic process noise $\boldsymbol{\Psi}$ be given by

$\begin{equation} \boldsymbol{\Psi} = \left( \begin{array}{cc} 0.25 & -0.1 \\ -0.1 & 0.22 \\ \end{array} \right) . \end{equation}$

R Function Arguments

n
#> [1] 100
time
#> [1] 10100
burnin
#> [1] 10000
# first mu0 in the list of length n
mu0[[1]]
#> [1] 0 0
# first sigma0 in the list of length n
sigma0[[1]]
#>            [,1]       [,2]
#> [1,]  0.3399277 -0.0557114
#> [2,] -0.0557114  1.1620385
# first sigma0_l in the list of length n
sigma0_l[[1]] # sigma0_l <- t(chol(sigma0))
#>             [,1]     [,2]
#> [1,]  0.58303323 0.000000
#> [2,] -0.09555442 1.073735
alpha
#> [[1]]
#> [1] 0 0
# first beta in the list of length n
beta[[1]]
#>           [,1]       [,2]
#> [1,] 0.5159622 0.01359474
#> [2,] 0.3202660 0.89904484
# first psi in the list of length n
psi[[1]]
#>       [,1]  [,2]
#> [1,]  0.25 -0.10
#> [2,] -0.10  0.22
psi_l[[1]] # psi_l <- t(chol(psi))
#>      [,1]      [,2]
#> [1,]  0.5 0.0000000
#> [2,] -0.2 0.4242641
# first nu in the list of length n
nu[[1]]
#> [1] 1.094299 0.100988
lambda
#> [[1]]
#>      [,1] [,2]
#> [1,]    1    0
#> [2,]    0    1
# first theta in the list of length n
theta[[1]]
#>      [,1] [,2]
#> [1,]  0.5  0.0
#> [2,]  0.0  0.5
theta_l[[1]] # theta_l <- t(chol(theta))
#>           [,1]      [,2]
#> [1,] 0.7071068 0.0000000
#> [2,] 0.0000000 0.7071068

Visualizing the Dynamics Without Process Noise and Measurement Error (n = 5 with Different Initial Condition)

Using the `SimSSMIVary` Function from the `simStateSpace` Package to Simulate Data

library(simStateSpace)
sim <- SimSSMIVary(
  n = n,
  time = time,
  mu0 = mu0,
  sigma0_l = sigma0_l,
  alpha = alpha,
  beta = beta,
  psi_l = psi_l,
  nu = nu,
  lambda = lambda,
  theta_l = theta_l
)
data <- as.data.frame(sim, burnin = burnin)
head(data)
#>   id time        y1         y2
#> 1  1    0 1.5903935 -1.0269628
#> 2  1    1 0.9288686  0.4296857
#> 3  1    2 2.0010930 -1.2365114
#> 4  1    3 2.2961315 -1.7734153
#> 5  1    4 1.9932917 -1.0469535
#> 6  1    5 1.4911727 -1.1539057
plot(sim, burnin = burnin)

Model Fitting

library(OpenMx)
library(fitDTVARMxID)

The FitDTVARMxID function fits a DT-VAR model on each individual $i$ . To set up the estimation, we first provide starting values for each parameter matrix.

Autoregressive Parameters (`beta`)

We initialize the autoregressive coefficient matrix $\boldsymbol{\beta}$ with the true values used in simulation.

beta_values <- beta

Intercepts (`nu`)

The intercept vector $\boldsymbol{\nu}$ is initialized with starting values.

nu_values <- nu

LDL’-parameterized covariance matrices

Covariances such as psi and theta are estimated using the LDL’ decomposition of a positive definite covariance matrix. The decomposition expresses a covariance matrix $\Sigma$ as

$\begin{equation} \boldsymbol{\Sigma} = \left( \mathbf{L} + \mathbf{I} \right) \mathrm{diag} \left( \mathrm{Softplus} \left( \mathbf{d}_{uc} \right) \right) \left( \mathbf{L} + \mathbf{I} \right)^{\prime}, \end{equation}$

where:

$\mathbf{L}$ is a strictly lower-triangular matrix of free parameters (l_mat_strict),
$\mathbf{I}$ is the identity matrix,
$\mathbf{d}_{uc}$ is an unconstrained vector,
$\mathrm{Softplus} \left(\mathbf{d}_{uc} \right) = \log \left(1 + \exp \left( \mathbf{d}_{uc} \right) \right)$ ensures strictly positive diagonal entries.

The LDL() function extracts this decomposition from a positive definite covariance matrix. It returns:

d_uc: unconstrained diagonal parameters, equal to InvSoftplus(d_vec),
d_vec: diagonal entries, equal to Softplus(d_uc),
l_mat_strict: the strictly lower-triangular factor.

sigma <- matrix(
  data = c(1.0, 0.5, 0.5, 1.0),
  nrow = 2,
  ncol = 2
)

ldl_sigma <- LDL(sigma)
d_uc <- ldl_sigma$d_uc
l_mat_strict <- ldl_sigma$l_mat_strict
I <- diag(2)
sigma_reconstructed <- (l_mat_strict + I) %*% diag(log1p(exp(d_uc)), 2) %*% t(l_mat_strict + I)
sigma_reconstructed
#>      [,1] [,2]
#> [1,]  1.0  0.5
#> [2,]  0.5  1.0

Process Noise Covariance Matrix (`psi`)

Starting values for the process noise covariance matrix $\boldsymbol{\Psi}$ are given below, with corresponding LDL’ parameters.

psi_values <- psi[[1]]
ldl_psi_values <- LDL(psi_values)
psi_d_values <- ldl_psi_values$d_uc
psi_l_values <- ldl_psi_values$l_mat_strict

psi_d_values
#> [1] -1.258692 -1.623449

psi_l_values
#>      [,1] [,2]
#> [1,]  0.0    0
#> [2,] -0.4    0

Measurement Error Covariance Matrix (`theta`)

Starting values for the measurement error covariance matrix $\boldsymbol{\Theta}$ are given below, with corresponding LDL’ parameters.

theta_values <- theta[[1]]
ldl_theta_values <- LDL(theta_values)
theta_d_values <- ldl_theta_values$d_uc
theta_l_values <- ldl_theta_values$l_mat_strict

theta_d_values
#> [1] -0.4327521 -0.4327521

theta_l_values
#>      [,1] [,2]
#> [1,]    0    0
#> [2,]    0    0

Initial mean vector (`mu_0`) and covariance matrix (`sigma_0`)

The initial mean vector $\boldsymbol{\mu_0}$ and covariance matrix $\boldsymbol{\Sigma_0}$ are fixed using mu0 and sigma0.

mu0_values <- mu0

sigma0_values <- lapply(
  X = sigma0,
  FUN = LDL
)
sigma0_d_values <- lapply(
  X = sigma0_values,
  FUN = function(i) {
    i$d_uc
  }
)
sigma0_l_values <- lapply(
  X = sigma0_values,
  FUN = function(i) {
    i$l_mat_strict
  }
)

`FitDTVARMxID`

fit <- FitDTVARMxID(
  data = data,
  observed = c("y1", "y2"),
  id = "id",
  beta_values = beta_values,
  psi_d_values = psi_d_values,
  psi_l_values = psi_l_values,
  nu_values = nu_values,
  theta_d_values = theta_d_values,
  mu0_values = mu0_values,
  sigma0_d_values = sigma0_d_values,
  sigma0_l_values = sigma0_l_values,
  ncores = parallel::detectCores()
)

Parameter estimates

summary(fit, converged = FALSE)
#> Call:
#> FitDTVARMxID(data = data, observed = c("y1", "y2"), id = "id", 
#>     beta_values = beta_values, psi_d_values = psi_d_values, psi_l_values = psi_l_values, 
#>     nu_values = nu_values, theta_d_values = theta_d_values, mu0_values = mu0_values, 
#>     sigma0_d_values = sigma0_d_values, sigma0_l_values = sigma0_l_values, 
#>     ncores = parallel::detectCores())
#> 
#> Estimated paramaters per individual.
#>                              beta_1_1 beta_2_1 beta_1_2 beta_2_2  nu_1_1
#> FitDTVARMxID_DTVAR_ID1.Rds     0.8919   0.3884   0.0014   0.8278  0.9393
#> FitDTVARMxID_DTVAR_ID2.Rds     0.5257   0.1873  -0.1914   0.8615  1.5692
#> FitDTVARMxID_DTVAR_ID3.Rds     0.4432   0.6642  -0.2631   0.7316  0.8984
#> FitDTVARMxID_DTVAR_ID4.Rds     0.5564   0.2990  -0.5390   1.0474  0.3122
#> FitDTVARMxID_DTVAR_ID5.Rds     0.4331   0.4058  -0.5188   1.2667  1.5956
#> FitDTVARMxID_DTVAR_ID6.Rds     0.7426   0.4695  -0.2859   0.7899  1.0874
#> FitDTVARMxID_DTVAR_ID7.Rds     0.0665   0.9187  -0.2204   1.0453  0.5087
#> FitDTVARMxID_DTVAR_ID8.Rds     0.8328  -0.4336   0.0099   0.3571  0.5714
#> FitDTVARMxID_DTVAR_ID9.Rds     0.6558   0.2997  -0.1429   0.7115  0.4182
#> FitDTVARMxID_DTVAR_ID10.Rds   -0.0888  -0.1026  -0.5923   0.7259  1.0406
#> FitDTVARMxID_DTVAR_ID11.Rds    0.5439   0.1095  -0.5159   0.7635  0.9207
#> FitDTVARMxID_DTVAR_ID12.Rds    0.7351   0.4251  -0.3198   0.7294  1.0544
#> FitDTVARMxID_DTVAR_ID14.Rds    0.5083   0.4353  -0.2188   0.5535  0.9219
#> FitDTVARMxID_DTVAR_ID15.Rds    0.5975   0.3000  -0.0495   0.1902  0.2497
#> FitDTVARMxID_DTVAR_ID16.Rds    0.4739   0.1934  -0.0865   0.8510  0.0967
#> FitDTVARMxID_DTVAR_ID17.Rds    0.6130   0.4445  -0.5719   0.8384  1.0589
#> FitDTVARMxID_DTVAR_ID18.Rds    0.3112   0.3257  -1.0749   0.9846  1.2933
#> FitDTVARMxID_DTVAR_ID19.Rds    0.6868   0.3807  -0.3537   0.4520 -0.0728
#> FitDTVARMxID_DTVAR_ID20.Rds    0.0637   0.1653  -0.4562   0.9036  1.0886
#> FitDTVARMxID_DTVAR_ID21.Rds    0.0308   0.1196  -0.8883   0.8563  0.7249
#> FitDTVARMxID_DTVAR_ID22.Rds    0.8479   0.4130  -0.4411   0.4283  0.3336
#> FitDTVARMxID_DTVAR_ID23.Rds    0.6044   1.0502   0.0576   0.5795  0.6397
#> FitDTVARMxID_DTVAR_ID24.Rds    0.3855   0.1433   0.0629   0.6544  0.9920
#> FitDTVARMxID_DTVAR_ID25.Rds    0.6035   0.8668   0.0219   0.2201  0.7772
#> FitDTVARMxID_DTVAR_ID26.Rds    0.7871   0.3098  -0.2562   0.7652  0.4026
#> FitDTVARMxID_DTVAR_ID27.Rds    0.7143   0.5289  -0.1746   0.4030  0.5877
#> FitDTVARMxID_DTVAR_ID28.Rds    0.7112   0.2936  -0.2609   0.8775  0.6417
#> FitDTVARMxID_DTVAR_ID29.Rds    0.7986   0.6522  -0.3993   0.6204  0.6727
#> FitDTVARMxID_DTVAR_ID30.Rds    0.6785   0.4018  -0.1128   0.8082  1.1349
#> FitDTVARMxID_DTVAR_ID31.Rds    0.7186  -0.1247   0.0573   0.1326  1.5822
#> FitDTVARMxID_DTVAR_ID32.Rds    0.4077   0.1000  -0.1997   0.5269  0.7587
#> FitDTVARMxID_DTVAR_ID33.Rds    0.0584  -0.1781  -0.2933   0.3247  0.6346
#> FitDTVARMxID_DTVAR_ID34.Rds    0.5858   0.3669  -0.2678   0.8685  1.0428
#> FitDTVARMxID_DTVAR_ID35.Rds    0.9933  -0.1423   0.0226   0.6535  0.9179
#> FitDTVARMxID_DTVAR_ID36.Rds    0.4150   0.1510  -0.2128   0.4070  1.2995
#> FitDTVARMxID_DTVAR_ID37.Rds    0.2994  -0.3453  -0.1303   0.5794  0.9056
#> FitDTVARMxID_DTVAR_ID38.Rds    0.8384   0.3829  -0.0509   0.7357  0.6339
#> FitDTVARMxID_DTVAR_ID39.Rds    0.6778   0.5315  -0.2441   0.9470 -0.4341
#> FitDTVARMxID_DTVAR_ID40.Rds    0.5687   0.2253  -0.3493   0.9319  0.7164
#> FitDTVARMxID_DTVAR_ID41.Rds    0.6643   0.1787  -0.4235   1.0045  0.6230
#> FitDTVARMxID_DTVAR_ID42.Rds    0.4616  -0.0774  -0.3266   0.4247  0.9917
#> FitDTVARMxID_DTVAR_ID43.Rds    0.4255   0.3454  -0.9733   0.8308  0.5968
#> FitDTVARMxID_DTVAR_ID44.Rds    0.7273  -0.3289   0.0232   0.1403  0.1864
#> FitDTVARMxID_DTVAR_ID45.Rds    0.2997   0.2276   0.0101   0.8528  1.2264
#> FitDTVARMxID_DTVAR_ID46.Rds    0.2213   0.5466  -0.6220   1.2917  1.0998
#> FitDTVARMxID_DTVAR_ID47.Rds    0.5446   0.2432  -0.4990   1.0842  0.9251
#> FitDTVARMxID_DTVAR_ID48.Rds    0.5364   0.2855  -0.5258   1.1674  0.8695
#> FitDTVARMxID_DTVAR_ID49.Rds    0.7391   0.1430  -0.2593   1.0413 -0.0530
#> FitDTVARMxID_DTVAR_ID50.Rds    0.5352   0.3243  -0.3550   0.8581  0.5771
#> FitDTVARMxID_DTVAR_ID51.Rds    0.5831   0.2484  -0.7171   0.7801 -0.2104
#> FitDTVARMxID_DTVAR_ID52.Rds    0.8203  -0.3590  -0.0564   0.2630  0.4539
#> FitDTVARMxID_DTVAR_ID54.Rds    0.6649   0.3054  -0.2069   0.8451  0.9380
#> FitDTVARMxID_DTVAR_ID55.Rds    0.9520  -0.3719   0.0223   0.0452  0.2008
#> FitDTVARMxID_DTVAR_ID56.Rds    0.7316   0.7897  -0.1380   0.1962 -0.0384
#> FitDTVARMxID_DTVAR_ID57.Rds    0.1999   0.2275  -2.4294   1.1564  1.1083
#> FitDTVARMxID_DTVAR_ID58.Rds   -0.1727   0.3624  -1.2892   1.2317  0.7826
#> FitDTVARMxID_DTVAR_ID59.Rds    0.1944   0.4585  -1.0374   0.7306  0.8993
#> FitDTVARMxID_DTVAR_ID60.Rds    0.2597   0.0173  -0.3095   0.7707  0.6641
#> FitDTVARMxID_DTVAR_ID61.Rds    0.3797   0.5476  -0.2347   1.0304  0.7983
#> FitDTVARMxID_DTVAR_ID62.Rds    0.6376   0.1371   0.1098   0.8688  0.8601
#> FitDTVARMxID_DTVAR_ID63.Rds    0.4156  -0.1558  -0.3166   0.2332  0.4532
#> FitDTVARMxID_DTVAR_ID64.Rds    0.6665   0.1510   0.1315   0.8073  1.2571
#> FitDTVARMxID_DTVAR_ID65.Rds    0.6053   0.2026  -0.8239   1.0986  1.2127
#> FitDTVARMxID_DTVAR_ID66.Rds    0.8298   0.3259  -0.2433   0.7246  0.3570
#> FitDTVARMxID_DTVAR_ID67.Rds    0.8597   0.5685  -0.2566   0.8274  0.5188
#> FitDTVARMxID_DTVAR_ID68.Rds    0.5832   0.1224  -0.3414   0.6805  0.6540
#> FitDTVARMxID_DTVAR_ID69.Rds    0.3752   0.0977  -2.0566   1.1563  0.2786
#> FitDTVARMxID_DTVAR_ID70.Rds    0.9673   0.5868  -0.2417   0.2854  1.4503
#> FitDTVARMxID_DTVAR_ID71.Rds    0.5242   0.3214  -0.2666   1.0273  1.4915
#> FitDTVARMxID_DTVAR_ID72.Rds    0.8269   0.1138  -0.5787   0.8983  1.3990
#> FitDTVARMxID_DTVAR_ID73.Rds    0.4556  -0.1167  -0.6133   0.2883  1.0499
#> FitDTVARMxID_DTVAR_ID74.Rds    0.8692   0.3334  -0.2986   0.5127  1.1073
#> FitDTVARMxID_DTVAR_ID75.Rds    0.5134   0.4089  -0.5192   0.9430  0.5005
#> FitDTVARMxID_DTVAR_ID76.Rds    0.6584  -0.0432  -0.2208  -0.1004  1.6525
#> FitDTVARMxID_DTVAR_ID77.Rds   -0.0031   0.5906  -0.8543   1.3146  0.6071
#> FitDTVARMxID_DTVAR_ID78.Rds    0.7896   0.1773  -0.0363  -0.1287  0.4759
#> FitDTVARMxID_DTVAR_ID79.Rds    0.5378   0.0052  -0.2816   0.1484  0.7595
#> FitDTVARMxID_DTVAR_ID81.Rds    0.4777   0.4558  -0.7610   0.7300  0.1346
#> FitDTVARMxID_DTVAR_ID82.Rds    0.2347   0.3121  -0.7700   0.8147  0.6759
#> FitDTVARMxID_DTVAR_ID83.Rds    0.5227   0.3757  -0.4201   1.0583  1.5295
#> FitDTVARMxID_DTVAR_ID84.Rds    0.3655   0.5335  -0.6958   1.1076  0.1993
#> FitDTVARMxID_DTVAR_ID85.Rds    0.4538   0.2503  -0.0661   0.8898  0.2049
#> FitDTVARMxID_DTVAR_ID86.Rds   -0.0170   0.0914  -0.5131   1.0274 -0.2605
#> FitDTVARMxID_DTVAR_ID87.Rds    0.1723   0.4414  -0.8348   0.9482  1.0610
#> FitDTVARMxID_DTVAR_ID88.Rds    0.8591  -0.0323  -0.0331   0.1639  0.9816
#> FitDTVARMxID_DTVAR_ID89.Rds    0.5429   0.4179  -0.0834   0.7373  0.8690
#> FitDTVARMxID_DTVAR_ID90.Rds    1.0423  -0.3218   0.1211   0.8422  1.6570
#> FitDTVARMxID_DTVAR_ID91.Rds    0.6725   0.0827  -0.2647   0.3960  1.3491
#> FitDTVARMxID_DTVAR_ID92.Rds    0.8519   0.0007  -0.1523   0.3439  1.0322
#> FitDTVARMxID_DTVAR_ID93.Rds    0.4460   0.3536  -0.7632   0.8995  0.7934
#> FitDTVARMxID_DTVAR_ID95.Rds    0.0203   0.0662  -0.0642   0.9206  0.7686
#> FitDTVARMxID_DTVAR_ID96.Rds    0.5754   0.0013  -0.2431   0.2766  0.0023
#> FitDTVARMxID_DTVAR_ID97.Rds    0.8955   0.3253  -0.4490   0.4666  1.2647
#> FitDTVARMxID_DTVAR_ID99.Rds    0.0016   1.0504  -0.7641   1.7562  0.2663
#> FitDTVARMxID_DTVAR_ID100.Rds   0.5632   0.2007  -0.1996   0.1181  0.8553
#>                               nu_2_1 psi_l_2_1 psi_d_1_1 psi_d_2_1 theta_d_1_1
#> FitDTVARMxID_DTVAR_ID1.Rds   -0.3761   -0.9968   -3.2463   -1.7312     -0.2036
#> FitDTVARMxID_DTVAR_ID2.Rds   -0.0221   -0.2510   -0.8351   -1.3210     -0.1664
#> FitDTVARMxID_DTVAR_ID3.Rds    0.3872   -0.4134   -1.3790   -3.2013     -0.4793
#> FitDTVARMxID_DTVAR_ID4.Rds    0.0387   -0.7876   -1.5329   -2.1633     -0.1318
#> FitDTVARMxID_DTVAR_ID5.Rds    0.0348   -0.4262   -1.4756  -23.6766     -0.3551
#> FitDTVARMxID_DTVAR_ID6.Rds    0.6085   -0.5134   -3.2397   -0.8957     -0.0399
#> FitDTVARMxID_DTVAR_ID7.Rds    0.1207   -0.4852   -0.9922   -7.7997     -0.3044
#> FitDTVARMxID_DTVAR_ID8.Rds    0.6513   -0.5290   -2.1272   -0.2610     -0.1714
#> FitDTVARMxID_DTVAR_ID9.Rds    0.3600   -0.4017   -0.6864   -2.9302     -0.6266
#> FitDTVARMxID_DTVAR_ID10.Rds   0.4899   -0.2600    0.5990   -1.5389    -16.7227
#> FitDTVARMxID_DTVAR_ID11.Rds   0.7276    0.0265    0.2256   -2.2004     -1.8771
#> FitDTVARMxID_DTVAR_ID12.Rds   0.5260   -0.2390   -0.6326   -2.0307     -0.7735
#> FitDTVARMxID_DTVAR_ID14.Rds   0.6606   -1.5986   -1.9585   -0.5001     -0.0245
#> FitDTVARMxID_DTVAR_ID15.Rds  -0.0856   -0.4318   -1.3139   -0.0434     -0.2197
#> FitDTVARMxID_DTVAR_ID16.Rds  -0.2623   -0.5343   -1.5818   -1.5964     -0.3221
#> FitDTVARMxID_DTVAR_ID17.Rds   0.8758   -0.5890   -2.4854   -3.0753      0.1470
#> FitDTVARMxID_DTVAR_ID18.Rds   1.1869    0.0468   -0.6264   -2.6657     -1.2760
#> FitDTVARMxID_DTVAR_ID19.Rds   0.2838   -0.1751   -0.5145   -2.3862     -0.5323
#> FitDTVARMxID_DTVAR_ID20.Rds   0.7181   -0.2176   -0.1320   -1.8417    -19.2365
#> FitDTVARMxID_DTVAR_ID21.Rds   0.2226    0.0273   -0.3272   -2.0091    -18.2531
#> FitDTVARMxID_DTVAR_ID22.Rds   0.3246   -5.6198   -4.1804  -18.7310     -0.1456
#> FitDTVARMxID_DTVAR_ID23.Rds   0.4575   -1.4625   -1.8543   -1.4715      0.1249
#> FitDTVARMxID_DTVAR_ID24.Rds   0.7390   -0.1243    0.0491   -1.7146    -20.5307
#> FitDTVARMxID_DTVAR_ID25.Rds   0.5366   -1.0748   -2.2854   -0.3640     -0.3860
#> FitDTVARMxID_DTVAR_ID26.Rds   0.4109   -0.3950   -0.5268   -1.2242     -1.0575
#> FitDTVARMxID_DTVAR_ID27.Rds   0.2882   -0.4157   -1.5882    0.2219     -0.3927
#> FitDTVARMxID_DTVAR_ID28.Rds   0.5094   -0.2771   -1.6365   -2.1678     -0.6743
#> FitDTVARMxID_DTVAR_ID29.Rds   0.5337   -0.8020   -2.6399   -1.3618     -0.3911
#> FitDTVARMxID_DTVAR_ID30.Rds   0.1974   -0.7215   -1.2781  -22.1749     -0.2214
#> FitDTVARMxID_DTVAR_ID31.Rds   0.6473   -0.2956   -1.6342   -0.3131     -0.3372
#> FitDTVARMxID_DTVAR_ID32.Rds   0.4787    0.0661   -0.1607   -1.2927     -3.0505
#> FitDTVARMxID_DTVAR_ID33.Rds   1.0122   -0.1023    0.3286   -0.1322    -21.5011
#> FitDTVARMxID_DTVAR_ID34.Rds   0.8781   -0.2366   -0.1554   -2.0352     -0.8470
#> FitDTVARMxID_DTVAR_ID35.Rds   0.7507   -4.6332   -3.5693  -21.9135     -0.2139
#> FitDTVARMxID_DTVAR_ID36.Rds   0.7965   -0.3183    0.0762   -0.3178     -2.1524
#> FitDTVARMxID_DTVAR_ID37.Rds   0.6572    0.1633   -1.2503   -0.2139     -0.5676
#> FitDTVARMxID_DTVAR_ID38.Rds   0.1413   -1.1470   -1.7593   -1.9109     -0.4197
#> FitDTVARMxID_DTVAR_ID39.Rds   0.2059   -1.0681   -2.7391   -3.5394     -0.1000
#> FitDTVARMxID_DTVAR_ID40.Rds   0.2540   -1.0250   -3.1138   -1.9371      0.1064
#> FitDTVARMxID_DTVAR_ID41.Rds   0.2476    1.0540   -2.9880  -22.5233     -0.3508
#> FitDTVARMxID_DTVAR_ID42.Rds   0.1485   -0.0390   -0.4160   -0.1322     -0.9683
#> FitDTVARMxID_DTVAR_ID43.Rds   0.1523   -1.2550   -2.6023  -21.6338     -0.5237
#> FitDTVARMxID_DTVAR_ID44.Rds   0.2816   -0.3165   -1.5336   -0.2952     -0.4791
#> FitDTVARMxID_DTVAR_ID45.Rds   0.9997   -0.4344   -0.6303   -2.3077     -1.0861
#> FitDTVARMxID_DTVAR_ID46.Rds   0.6119   -0.1538   -1.4753   -4.6425     -0.7591
#> FitDTVARMxID_DTVAR_ID47.Rds   0.0259   -0.3535   -1.0740   -4.4203     -0.1339
#> FitDTVARMxID_DTVAR_ID48.Rds   1.1838   -0.3531   -1.0703   -5.1402     -0.9649
#> FitDTVARMxID_DTVAR_ID49.Rds   0.9898   -0.4070   -1.8896  -25.2097     -0.0450
#> FitDTVARMxID_DTVAR_ID50.Rds   0.6829   -0.6733   -1.1095  -22.3107     -0.6209
#> FitDTVARMxID_DTVAR_ID51.Rds  -0.2341    0.0704   -0.7940   -2.0126     -0.7923
#> FitDTVARMxID_DTVAR_ID52.Rds   0.4256   -0.4046   -1.4416    0.4011     -0.7487
#> FitDTVARMxID_DTVAR_ID54.Rds   0.6324   -0.1045   -1.0674   -2.2570     -0.9457
#> FitDTVARMxID_DTVAR_ID55.Rds   0.5551   -2.5275   -2.4432   -2.8985     -0.4571
#> FitDTVARMxID_DTVAR_ID56.Rds   0.1693   -1.6237   -1.4744  -18.9790     -0.5271
#> FitDTVARMxID_DTVAR_ID57.Rds   0.7444    1.1140   -4.7221  -17.0145     -0.4377
#> FitDTVARMxID_DTVAR_ID58.Rds   0.1264    0.0723   -0.8839  -20.8778     -0.9791
#> FitDTVARMxID_DTVAR_ID59.Rds   0.1729   -0.6768   -1.7563   -2.7574     -0.3852
#> FitDTVARMxID_DTVAR_ID60.Rds   0.3725   -0.1195   -0.0582   -0.9027    -20.6515
#> FitDTVARMxID_DTVAR_ID61.Rds   0.1216    0.1334   -1.5613  -18.3586     -0.5007
#> FitDTVARMxID_DTVAR_ID62.Rds   1.2482   -0.3308   -0.5665   -3.5576     -1.4692
#> FitDTVARMxID_DTVAR_ID63.Rds   1.2683   -0.1339   -0.2828    0.1342     -0.8319
#> FitDTVARMxID_DTVAR_ID64.Rds   0.1519   -1.4533   -2.2546   -4.6968     -0.4356
#> FitDTVARMxID_DTVAR_ID65.Rds   0.4718    0.1292   -1.7872  -23.2366     -0.6194
#> FitDTVARMxID_DTVAR_ID66.Rds   0.7091   -0.1220   -0.8826   -1.0849     -0.1779
#> FitDTVARMxID_DTVAR_ID67.Rds   0.1261   -1.0181   -1.4179   -2.6693     -0.4463
#> FitDTVARMxID_DTVAR_ID68.Rds   0.3936   -0.3996   -1.4347   -1.1631     -0.7953
#> FitDTVARMxID_DTVAR_ID69.Rds   0.4710    0.1331   -1.5910  -20.5082     -0.8913
#> FitDTVARMxID_DTVAR_ID70.Rds   0.7958   -2.1605   -2.3243   -1.6170     -0.1558
#> FitDTVARMxID_DTVAR_ID71.Rds   0.6112   -0.4584   -0.8158  -24.0349     -0.7822
#> FitDTVARMxID_DTVAR_ID72.Rds   0.7881    0.0940   -2.9338   -5.1466     -0.1564
#> FitDTVARMxID_DTVAR_ID73.Rds   0.1736   -1.2758   -2.3210   -1.6518     -0.4570
#> FitDTVARMxID_DTVAR_ID74.Rds   0.2534   -0.8966   -1.1497  -19.0540     -0.2441
#> FitDTVARMxID_DTVAR_ID75.Rds   0.3778   -0.6101   -1.2912   -2.5382     -0.7904
#> FitDTVARMxID_DTVAR_ID76.Rds   0.3116   -0.0590   -1.3322   -0.2085     -0.8085
#> FitDTVARMxID_DTVAR_ID77.Rds   0.9099   -0.4509   -0.7086   -4.2806     -1.0550
#> FitDTVARMxID_DTVAR_ID78.Rds   1.1462   -0.8064   -2.0103   -0.2479     -0.2994
#> FitDTVARMxID_DTVAR_ID79.Rds   0.4002   -0.1145   -0.4128    0.0694     -2.1749
#> FitDTVARMxID_DTVAR_ID81.Rds   0.5828   -2.8204   -3.7344  -20.2643     -0.0762
#> FitDTVARMxID_DTVAR_ID82.Rds   0.4393   -0.2260   -0.7880   -2.2414     -1.4845
#> FitDTVARMxID_DTVAR_ID83.Rds   0.9268   -0.6248   -1.5440  -23.9969     -0.0284
#> FitDTVARMxID_DTVAR_ID84.Rds   0.6800   -0.3656   -2.5514   -2.7434     -0.5485
#> FitDTVARMxID_DTVAR_ID85.Rds   0.9179   -0.3722   -1.1764   -2.3544     -0.8277
#> FitDTVARMxID_DTVAR_ID86.Rds   1.6928   -0.0466    0.2346   -2.3618    -17.4284
#> FitDTVARMxID_DTVAR_ID87.Rds   0.2380   -0.1283   -0.7710   -2.5302     -1.0472
#> FitDTVARMxID_DTVAR_ID88.Rds   0.4145   -1.1653   -1.8173   -0.0499     -0.4420
#> FitDTVARMxID_DTVAR_ID89.Rds   0.9013   -0.5322   -0.8380   -1.1527     -0.7745
#> FitDTVARMxID_DTVAR_ID90.Rds   0.6951    1.3942   -4.8391  -22.1085     -0.0465
#> FitDTVARMxID_DTVAR_ID91.Rds   0.9365   -1.1931   -1.6550   -0.0975     -0.4430
#> FitDTVARMxID_DTVAR_ID92.Rds   1.0531   -0.7311   -2.9361    0.3114      0.0418
#> FitDTVARMxID_DTVAR_ID93.Rds  -0.1439   -0.7725   -2.4819   -2.4791     -0.6264
#> FitDTVARMxID_DTVAR_ID95.Rds   0.3106   -0.1162    0.2415   -1.9212    -16.6985
#> FitDTVARMxID_DTVAR_ID96.Rds   0.2766   -0.0109   -1.0388   -0.1553     -0.8936
#> FitDTVARMxID_DTVAR_ID97.Rds   0.8966   -0.8683   -2.3775   -0.6068     -0.4797
#> FitDTVARMxID_DTVAR_ID99.Rds   0.5161   -0.7345   -1.8743  -22.0498     -0.1484
#> FitDTVARMxID_DTVAR_ID100.Rds  0.3247   -0.3649   -0.6394    0.0362     -0.7331
#>                              theta_d_2_1
#> FitDTVARMxID_DTVAR_ID1.Rds       -0.6692
#> FitDTVARMxID_DTVAR_ID2.Rds       -0.1704
#> FitDTVARMxID_DTVAR_ID3.Rds        0.1474
#> FitDTVARMxID_DTVAR_ID4.Rds       -0.9299
#> FitDTVARMxID_DTVAR_ID5.Rds        0.3114
#> FitDTVARMxID_DTVAR_ID6.Rds       -0.7680
#> FitDTVARMxID_DTVAR_ID7.Rds       -0.3588
#> FitDTVARMxID_DTVAR_ID8.Rds       -1.4516
#> FitDTVARMxID_DTVAR_ID9.Rds       -0.0424
#> FitDTVARMxID_DTVAR_ID10.Rds      -0.4068
#> FitDTVARMxID_DTVAR_ID11.Rds      -0.3336
#> FitDTVARMxID_DTVAR_ID12.Rds      -0.2473
#> FitDTVARMxID_DTVAR_ID14.Rds     -18.1336
#> FitDTVARMxID_DTVAR_ID15.Rds     -19.7042
#> FitDTVARMxID_DTVAR_ID16.Rds      -0.6846
#> FitDTVARMxID_DTVAR_ID17.Rds      -0.6269
#> FitDTVARMxID_DTVAR_ID18.Rds      -0.4967
#> FitDTVARMxID_DTVAR_ID19.Rds      -0.6452
#> FitDTVARMxID_DTVAR_ID20.Rds      -0.3319
#> FitDTVARMxID_DTVAR_ID21.Rds      -0.2220
#> FitDTVARMxID_DTVAR_ID22.Rds      -1.1265
#> FitDTVARMxID_DTVAR_ID23.Rds      -2.3342
#> FitDTVARMxID_DTVAR_ID24.Rds      -0.5339
#> FitDTVARMxID_DTVAR_ID25.Rds      -3.7317
#> FitDTVARMxID_DTVAR_ID26.Rds      -0.8024
#> FitDTVARMxID_DTVAR_ID27.Rds     -19.6656
#> FitDTVARMxID_DTVAR_ID28.Rds      -0.3447
#> FitDTVARMxID_DTVAR_ID29.Rds      -0.5756
#> FitDTVARMxID_DTVAR_ID30.Rds      -0.3066
#> FitDTVARMxID_DTVAR_ID31.Rds     -19.9108
#> FitDTVARMxID_DTVAR_ID32.Rds      -1.0550
#> FitDTVARMxID_DTVAR_ID33.Rds     -27.8688
#> FitDTVARMxID_DTVAR_ID34.Rds      -0.7411
#> FitDTVARMxID_DTVAR_ID35.Rds      -1.9449
#> FitDTVARMxID_DTVAR_ID36.Rds      -2.5229
#> FitDTVARMxID_DTVAR_ID37.Rds      -0.8266
#> FitDTVARMxID_DTVAR_ID38.Rds      -0.7667
#> FitDTVARMxID_DTVAR_ID39.Rds      -0.3469
#> FitDTVARMxID_DTVAR_ID40.Rds      -0.5945
#> FitDTVARMxID_DTVAR_ID41.Rds      -0.2828
#> FitDTVARMxID_DTVAR_ID42.Rds      -1.0004
#> FitDTVARMxID_DTVAR_ID43.Rds      -0.4565
#> FitDTVARMxID_DTVAR_ID44.Rds     -17.0464
#> FitDTVARMxID_DTVAR_ID45.Rds      -0.2269
#> FitDTVARMxID_DTVAR_ID46.Rds       0.2719
#> FitDTVARMxID_DTVAR_ID47.Rds      -0.2751
#> FitDTVARMxID_DTVAR_ID48.Rds       0.1380
#> FitDTVARMxID_DTVAR_ID49.Rds       0.0072
#> FitDTVARMxID_DTVAR_ID50.Rds      -0.2316
#> FitDTVARMxID_DTVAR_ID51.Rds      -0.5527
#> FitDTVARMxID_DTVAR_ID52.Rds     -18.1112
#> FitDTVARMxID_DTVAR_ID54.Rds      -0.5366
#> FitDTVARMxID_DTVAR_ID55.Rds      -1.4157
#> FitDTVARMxID_DTVAR_ID56.Rds      -1.0502
#> FitDTVARMxID_DTVAR_ID57.Rds       0.0530
#> FitDTVARMxID_DTVAR_ID58.Rds      -0.2894
#> FitDTVARMxID_DTVAR_ID59.Rds      -0.6244
#> FitDTVARMxID_DTVAR_ID60.Rds      -0.7666
#> FitDTVARMxID_DTVAR_ID61.Rds      -0.0504
#> FitDTVARMxID_DTVAR_ID62.Rds      -0.3200
#> FitDTVARMxID_DTVAR_ID63.Rds     -19.0645
#> FitDTVARMxID_DTVAR_ID64.Rds      -0.0648
#> FitDTVARMxID_DTVAR_ID65.Rds      -0.1107
#> FitDTVARMxID_DTVAR_ID66.Rds      -1.0690
#> FitDTVARMxID_DTVAR_ID67.Rds      -0.7145
#> FitDTVARMxID_DTVAR_ID68.Rds      -0.7932
#> FitDTVARMxID_DTVAR_ID69.Rds      -0.0205
#> FitDTVARMxID_DTVAR_ID70.Rds     -20.3551
#> FitDTVARMxID_DTVAR_ID71.Rds      -0.1750
#> FitDTVARMxID_DTVAR_ID72.Rds      -0.4325
#> FitDTVARMxID_DTVAR_ID73.Rds      -0.7988
#> FitDTVARMxID_DTVAR_ID74.Rds      -0.2854
#> FitDTVARMxID_DTVAR_ID75.Rds      -0.3703
#> FitDTVARMxID_DTVAR_ID76.Rds     -19.1391
#> FitDTVARMxID_DTVAR_ID77.Rds      -0.0015
#> FitDTVARMxID_DTVAR_ID78.Rds     -19.0998
#> FitDTVARMxID_DTVAR_ID79.Rds     -17.3563
#> FitDTVARMxID_DTVAR_ID81.Rds      -0.5221
#> FitDTVARMxID_DTVAR_ID82.Rds      -0.4074
#> FitDTVARMxID_DTVAR_ID83.Rds      -0.4105
#> FitDTVARMxID_DTVAR_ID84.Rds      -0.2260
#> FitDTVARMxID_DTVAR_ID85.Rds      -0.3929
#> FitDTVARMxID_DTVAR_ID86.Rds      -0.1352
#> FitDTVARMxID_DTVAR_ID87.Rds      -0.6746
#> FitDTVARMxID_DTVAR_ID88.Rds     -16.3882
#> FitDTVARMxID_DTVAR_ID89.Rds      -0.1831
#> FitDTVARMxID_DTVAR_ID90.Rds       0.1473
#> FitDTVARMxID_DTVAR_ID91.Rds     -20.6057
#> FitDTVARMxID_DTVAR_ID92.Rds     -20.4702
#> FitDTVARMxID_DTVAR_ID93.Rds      -0.3766
#> FitDTVARMxID_DTVAR_ID95.Rds      -0.4413
#> FitDTVARMxID_DTVAR_ID96.Rds     -19.2237
#> FitDTVARMxID_DTVAR_ID97.Rds      -1.0194
#> FitDTVARMxID_DTVAR_ID99.Rds      -0.3245
#> FitDTVARMxID_DTVAR_ID100.Rds    -18.3309

Proportion of converged cases

converged(
  fit,
  theta_tol = 0.01,
  prop = TRUE
)
#> [1] 0.71

Fixed-Effect Meta-Analysis of Measurement Error

When fitting DT-VAR models per person, separating process noise ( $\boldsymbol{\Psi}$ ) from measurement error ( $\boldsymbol{\Theta}$ ) can be unstable for some individuals. To stabilize inference, we first pool the person-level $\boldsymbol{\Theta}_{i}$ estimates from only the converged fits using a fixed-effect meta-analysis. This yields a high-precision estimate of the common measurement-error covariance that we will then hold fixed in a second pass of model fitting.

What the code does: - Selects individuals that converged and whose $\boldsymbol{\Theta}_i$ diagonals exceed a small threshold (theta_tol), filtering out near-zero or ill-conditioned solutions. - Extracts each person’s LDL’ diagonal parameters for $\boldsymbol{\Theta}_i$ and their sampling covariance matrices. - Computes the inverse-variance-weighted pooled estimate (fixed effect), returning it on the same LDL’ parameterization used by FitDTVARMxID().

library(metaVAR)
fixed_theta <- MetaVARMx(
  fit,
  random = FALSE, # TRUE by default
  effects = FALSE, # TRUE by default
  cov_meas = TRUE, # FALSE by default
  theta_tol = 0.01
)

You can read summary(fixed_theta) as providing the pooled (fixed) measurement-error scale that is common across persons. If individual instruments truly share the same reliability structure, fixing $\boldsymbol{\Theta}$ to this pooled value improves stability and often reduces bias in the dynamic parameters.

Note: Fixed-effect pooling assumes a common $\boldsymbol{\Theta}$ across individuals.

coef(fixed_theta)
#>  alpha_1_1  alpha_2_1 
#> -0.3458380 -0.3159435
summary(fixed_theta)
#> Call:
#> MetaVARMx(object = fit, random = FALSE, effects = FALSE, cov_meas = TRUE, 
#>     theta_tol = 0.01)
#>                est     se       z p    2.5%   97.5%
#> alpha[1,1] -0.3458 0.0397 -8.7020 0 -0.4237 -0.2679
#> alpha[2,1] -0.3159 0.0357 -8.8425 0 -0.3860 -0.2459

theta_d_values <- coef(fixed_theta)

Refit the model with fixed measurement error covariance matrix

We refit the individual models using the pooled $\boldsymbol{\Theta}$ as a fixed measurement-error covariance matrix.

fit <- FitDTVARMxID(
  data = data,
  observed = c("y1", "y2"),
  id = "id",
  beta_values = beta_values,
  psi_d_values = psi_d_values,
  psi_l_values = psi_l_values,
  nu_values = nu_values,
  theta_fixed = TRUE,
  theta_d_values = theta_d_values,
  mu0_values = mu0_values,
  sigma0_d_values = sigma0_d_values,
  sigma0_l_values = sigma0_l_values,
  ncores = parallel::detectCores()
)

With $\boldsymbol{\Theta}$ fixed, the re-estimation focuses on the dynamic structure ( $\boldsymbol{\beta}$ , $\boldsymbol{\Psi}$ ) and intercepts ( $\boldsymbol{\nu}$ ). In practice, this often increases the proportion of converged fits and yields more stable cross-lag estimates.

Proportion of converged cases

converged(
  fit,
  prop = TRUE
)
#> [1] 0.99

Random-Effects Meta-Analysis of Person-Specific Dynamics and Means

Having stabilized $\boldsymbol{\Theta}$ , we synthesize the person-specific estimates to recover population-level effects and their between-person variability. We use a random-effects model so the pooled mean reflects both within-person estimation uncertainty and between-person heterogeneity.

random <- MetaVARMx(
  fit,
  effects = TRUE,
  int_meas = TRUE
)

summary(random)
#> Call:
#> MetaVARMx(object = fit, effects = TRUE, int_meas = TRUE)
#>                  est     se        z      p    2.5%   97.5%
#> alpha[1,1]    0.6940 0.0193  35.9431 0.0000  0.6562  0.7319
#> alpha[2,1]    0.2686 0.0175  15.3185 0.0000  0.2342  0.3029
#> alpha[3,1]   -0.2463 0.0266  -9.2487 0.0000 -0.2985 -0.1941
#> alpha[4,1]    0.8305 0.0209  39.7511 0.0000  0.7896  0.8715
#> alpha[5,1]    0.7585 0.0454  16.7138 0.0000  0.6695  0.8474
#> alpha[6,1]    0.5035 0.0345  14.5931 0.0000  0.4359  0.5712
#> tau_sqr[1,1]  0.0085 0.0037   2.3029 0.0213  0.0013  0.0158
#> tau_sqr[2,1]  0.0031 0.0026   1.1867 0.2354 -0.0020  0.0081
#> tau_sqr[3,1]  0.0035 0.0040   0.8809 0.3784 -0.0043  0.0113
#> tau_sqr[4,1] -0.0037 0.0029  -1.2558 0.2092 -0.0094  0.0021
#> tau_sqr[5,1] -0.0008 0.0075  -0.1036 0.9175 -0.0156  0.0140
#> tau_sqr[6,1]  0.0032 0.0056   0.5609 0.5748 -0.0079  0.0142
#> tau_sqr[2,2]  0.0094 0.0031   3.0150 0.0026  0.0033  0.0155
#> tau_sqr[3,2] -0.0046 0.0037  -1.2389 0.2154 -0.0119  0.0027
#> tau_sqr[4,2] -0.0059 0.0027  -2.1578 0.0309 -0.0112 -0.0005
#> tau_sqr[5,2] -0.0034 0.0068  -0.4988 0.6179 -0.0167  0.0099
#> tau_sqr[6,2] -0.0046 0.0051  -0.9071 0.3643 -0.0147  0.0054
#> tau_sqr[3,3]  0.0266 0.0076   3.4999 0.0005  0.0117  0.0415
#> tau_sqr[4,3]  0.0048 0.0043   1.1305 0.2583 -0.0035  0.0131
#> tau_sqr[5,3]  0.0103 0.0106   0.9750 0.3296 -0.0104  0.0311
#> tau_sqr[6,3]  0.0123 0.0087   1.4218 0.1551 -0.0047  0.0293
#> tau_sqr[4,4]  0.0158 0.0053   2.9729 0.0030  0.0054  0.0261
#> tau_sqr[5,4]  0.0034 0.0085   0.3964 0.6918 -0.0133  0.0201
#> tau_sqr[6,4] -0.0042 0.0063  -0.6664 0.5051 -0.0167  0.0082
#> tau_sqr[5,5]  0.1828 0.0286   6.3986 0.0000  0.1268  0.2388
#> tau_sqr[6,5]  0.0454 0.0158   2.8764 0.0040  0.0145  0.0764
#> tau_sqr[6,6]  0.0960 0.0162   5.9143 0.0000  0.0642  0.1278
#> i_sqr[1,1]    0.4921 0.1085   4.5345 0.0000  0.2794  0.7048
#> i_sqr[2,1]    0.6871 0.0603  11.3892 0.0000  0.5689  0.8054
#> i_sqr[3,1]    0.8113 0.0385  21.0629 0.0000  0.7358  0.8868
#> i_sqr[4,1]    0.9139 0.0167  54.8647 0.0000  0.8813  0.9466
#> i_sqr[5,1]    0.9362 0.0093 100.2964 0.0000  0.9179  0.9545
#> i_sqr[6,1]    0.9137 0.0130  70.1295 0.0000  0.8882  0.9392

The fixed part of the random-effects model gives pooled means $\boldsymbol{\alpha} = \mathbb{E} \left[ \mathrm{Vec} \left( \boldsymbol{\beta} \right), \boldsymbol{\nu} \right]$ .
The random part yields between-person covariances ( $\boldsymbol{\tau}^{2}$ ) quantifying heterogeneity in dynamics ( $\boldsymbol{\beta}$ ) and means ( $\boldsymbol{\nu}$ ) across individuals.

means <- extract(random, what = "alpha")
means
#> [1]  0.6940078  0.2685507 -0.2463056  0.8305160  0.7584888  0.5035365
covariances <- extract(random, what = "tau_sqr")
covariances
#>               [,1]         [,2]         [,3]         [,4]          [,5]
#> [1,]  0.0085125608  0.003055081  0.003507298 -0.003684972 -0.0007809232
#> [2,]  0.0030550808  0.009397715 -0.004593901 -0.005892504 -0.0033961601
#> [3,]  0.0035072977 -0.004593901  0.026590461  0.004808733  0.0103399611
#> [4,] -0.0036849718 -0.005892504  0.004808733  0.015758679  0.0033753141
#> [5,] -0.0007809232 -0.003396160  0.010339961  0.003375314  0.1828274508
#> [6,]  0.0031655041 -0.004648634  0.012332324 -0.004227459  0.0454136666
#>              [,6]
#> [1,]  0.003165504
#> [2,] -0.004648634
#> [3,]  0.012332324
#> [4,] -0.004227459
#> [5,]  0.045413667
#> [6,]  0.095988254

Finally, we compare the meta-analytic population estimates to the known generating values.

beta_pop_mean
#> [1]  0.5975226  0.2495715 -0.3054649  0.6893696
beta_pop_cov
#>               [,1]         [,2]         [,3]          [,4]
#> [1,]  2.386517e-02 0.0092591880 8.078956e-05 -0.0002817067
#> [2,]  9.259188e-03 0.0193540268 1.323752e-04  0.0001410741
#> [3,]  8.078956e-05 0.0001323752 1.883487e-02  0.0084724601
#> [4,] -2.817067e-04 0.0001410741 8.472460e-03  0.0222254212

nu_mu
#> [1] 0.8 0.5
nu_sigma
#>      [,1] [,2]
#> [1,] 0.15 0.05
#> [2,] 0.05 0.10

Summary

This vignette demonstrates a two-stage hierarchical estimation approach for dynamic systems: 1. individual-level DT-VAR estimation with stabilized measurement error, and
2. population-level meta-analysis of person-specific dynamics and means.

References

Hunter, M. D. (2017). State space modeling in an open source, modular, structural equation modeling environment. Structural Equation Modeling: A Multidisciplinary Journal, 25(2), 307–324. https://doi.org/10.1080/10705511.2017.1369354

Neale, M. C., Hunter, M. D., Pritikin, J. N., Zahery, M., Brick, T. R., Kirkpatrick, R. M., Estabrook, R., Bates, T. C., Maes, H. H., & Boker, S. M. (2015). OpenMx 2.0: Extended structural equation and statistical modeling. Psychometrika, 81(2), 535–549. https://doi.org/10.1007/s11336-014-9435-8

R Core Team. (2024). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/

Fit the Discrete-Time Vector Autoregressive Model By ID (Adaptive Recovery | 100 Measurement Occasions)

Ivan Jacob Agaloos Pesigan

2025-10-22

Dynamics Description

Model

Data Generation

Notation

R Function Arguments

Visualizing the Dynamics Without Process Noise and Measurement Error (n = 5 with Different Initial Condition)

Using the `SimSSMIVary` Function from the `simStateSpace` Package to Simulate Data

Model Fitting

Autoregressive Parameters (`beta`)

Intercepts (`nu`)

LDL’-parameterized covariance matrices

Process Noise Covariance Matrix (`psi`)

Measurement Error Covariance Matrix (`theta`)

Initial mean vector (`mu_0`) and covariance matrix (`sigma_0`)

`FitDTVARMxID`

Parameter estimates

Proportion of converged cases

Fixed-Effect Meta-Analysis of Measurement Error

Refit the model with fixed measurement error covariance matrix

Proportion of converged cases

Random-Effects Meta-Analysis of Person-Specific Dynamics and Means

Summary

References

Fit the Discrete-Time Vector Autoregressive Model By ID (Adaptive Recovery | 100 Measurement Occasions)

Ivan Jacob Agaloos Pesigan

2025-10-22

Dynamics Description

Model

Data Generation

Notation

R Function Arguments

Visualizing the Dynamics Without Process Noise and Measurement Error (n = 5 with Different Initial Condition)

Using the SimSSMIVary Function from the simStateSpace Package to Simulate Data

Model Fitting

Autoregressive Parameters (beta)

Intercepts (nu)

LDL’-parameterized covariance matrices

Process Noise Covariance Matrix (psi)

Measurement Error Covariance Matrix (theta)

Initial mean vector (mu_0) and covariance matrix (sigma_0)

FitDTVARMxID

Parameter estimates

Proportion of converged cases

Fixed-Effect Meta-Analysis of Measurement Error

Refit the model with fixed measurement error covariance matrix

Proportion of converged cases

Random-Effects Meta-Analysis of Person-Specific Dynamics and Means

Summary

References

Using the `SimSSMIVary` Function from the `simStateSpace` Package to Simulate Data

Autoregressive Parameters (`beta`)

Intercepts (`nu`)

Process Noise Covariance Matrix (`psi`)

Measurement Error Covariance Matrix (`theta`)

Initial mean vector (`mu_0`) and covariance matrix (`sigma_0`)

`FitDTVARMxID`