Apply a logistic normal multinomial model to jointly model a vector of count responses $y$ in an outcome or mediation model. This is a common choice for data where the parameter of interest is the composition across responses (e.g., microbiome).
Value
model An object of class model with estimator, predictor, and
sampler functions associated wtih a linear model.
Examples
m <- lnm_model()
mat <- data.frame(matrix(rpois(250, 10), 25, 10))
colnames(mat) <- paste0("y", seq_len(6))
fit <- estimator(m)(y1 + y2 + y3 + y4 ~ y5 + y6, mat)
#> Chain 1: ------------------------------------------------------------
#> Chain 1: EXPERIMENTAL ALGORITHM:
#> Chain 1: This procedure has not been thoroughly tested and may be unstable
#> Chain 1: or buggy. The interface is subject to change.
#> Chain 1: ------------------------------------------------------------
#> Chain 1:
#> Chain 1:
#> Chain 1:
#> Chain 1: Gradient evaluation took 9e-05 seconds
#> Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 0.9 seconds.
#> Chain 1: Adjust your expectations accordingly!
#> Chain 1:
#> Chain 1:
#> Chain 1: Begin eta adaptation.
#> Chain 1: Iteration: 1 / 250 [ 0%] (Adaptation)
#> Chain 1: Iteration: 50 / 250 [ 20%] (Adaptation)
#> Chain 1: Iteration: 100 / 250 [ 40%] (Adaptation)
#> Chain 1: Iteration: 150 / 250 [ 60%] (Adaptation)
#> Chain 1: Iteration: 200 / 250 [ 80%] (Adaptation)
#> Chain 1: Success! Found best value [eta = 1] earlier than expected.
#> Chain 1:
#> Chain 1: Begin stochastic gradient ascent.
#> Chain 1: iter ELBO delta_ELBO_mean delta_ELBO_med notes
#> Chain 1: 100 -5574.268 1.000 1.000
#> Chain 1: 200 -350.516 7.952 14.903
#> Chain 1: 300 -293.670 5.366 1.000
#> Chain 1: 400 -255.233 4.062 1.000
#> Chain 1: 500 -248.825 3.255 0.194
#> Chain 1: 600 -262.916 2.721 0.194
#> Chain 1: 700 -250.397 2.340 0.151
#> Chain 1: 800 -256.953 2.050 0.151
#> Chain 1: 900 -283.830 1.833 0.095
#> Chain 1: 1000 -245.215 1.665 0.151
#> Chain 1: 1100 -244.821 1.566 0.095 MAY BE DIVERGING... INSPECT ELBO
#> Chain 1: 1200 -248.446 0.077 0.054
#> Chain 1: 1300 -265.950 0.064 0.054
#> Chain 1: 1400 -263.389 0.050 0.050
#> Chain 1: 1500 -257.201 0.050 0.050
#> Chain 1: 1600 -247.242 0.048 0.040
#> Chain 1: 1700 -242.917 0.045 0.026
#> Chain 1: 1800 -245.240 0.044 0.024
#> Chain 1: 1900 -242.659 0.035 0.018
#> Chain 1: 2000 -241.643 0.020 0.015
#> Chain 1: 2100 -243.343 0.020 0.015
#> Chain 1: 2200 -250.057 0.022 0.018
#> Chain 1: 2300 -246.020 0.017 0.016
#> Chain 1: 2400 -246.665 0.016 0.016
#> Chain 1: 2500 -240.445 0.016 0.016
#> Chain 1: 2600 -240.664 0.012 0.011
#> Chain 1: 2700 -242.534 0.011 0.009 MEDIAN ELBO CONVERGED
#> Chain 1:
#> Chain 1: Drawing a sample of size 1000 from the approximate posterior...
#> Chain 1: COMPLETED.
#> Warning: Pareto k diagnostic value is 1.29. Resampling is disabled. Decreasing tol_rel_obj may help if variational algorithm has terminated prematurely. Otherwise consider using sampling instead.