GEV responses#

In this tutorial, we illustrate how to set up a distributional regression model with the generalized extreme value distribution as a response distribution. First, we simulate some data in R:

  • The location parameter (\(\mu\)) is a function of an intercept and a non-linear covariate effect.

  • The scale parameter (\(\sigma\)) is a function of an intercept and a linear effect and uses a log-link.

  • The shape or concentration parameter (\(\xi\)) is a function of an intercept and a linear effect.

After simulating the data, we can configure the model with a single call to the rliesel::liesel() function.

library(rliesel)

Please set your Liesel venv, e.g. with use_liesel_venv()

library(VGAM)

Loading required package: stats4

Loading required package: splines

set.seed(1337)

n <- 1000

x0 <- runif(n)
x1 <- runif(n)
x2 <- runif(n)

y <- rgev(
  n,
  location = 0 + sin(2 * pi * x0),
  scale = exp(-3 + x1),
  shape = 0.1 + x2
)

plot(y)


model <- liesel(
  response = y,
  distribution = "GeneralizedExtremeValue",
  predictors = list(
    loc = predictor(~ s(x0)),
    scale = predictor(~ x1, inverse_link = "Exp"),
    concentration = predictor(~ x2)
  )
)

Now, we can continue in Python and use the lsl.dist_reg_mcmc() function to set up a sampling algorithm with IWLS kernels for the regression coefficients (\(\boldsymbol{\beta}\)) and a Gibbs kernel for the smoothing parameter (\(\tau^2\)) of the spline. Note that we need to set \(\beta_0\) for \(\xi\) to 0.1 manually, because \(\xi = 0\) breaks the sampler.

import liesel.model as lsl
import jax.numpy as jnp

model = r.model

# concentration == 0.0 seems to break the sampler
model.vars["concentration_p0_beta"].value = jnp.array([0.1, 0.0])

builder = lsl.dist_reg_mcmc(model, seed=42, num_chains=4)
builder.set_duration(warmup_duration=1000, posterior_duration=1000)

engine = builder.build()
engine.sample_all_epochs()

liesel.goose.engine - INFO - Starting epoch: FAST_ADAPTATION, 75 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 8, 11, 5, 4 / 75 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 0, 0, 1, 0 / 75 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 1, 0, 0, 0 / 75 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Starting epoch: SLOW_ADAPTATION, 25 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 0, 1, 2, 3 / 25 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 1, 1, 1, 1 / 25 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 1, 1, 1, 1 / 25 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_04: 2, 1, 3, 1 / 25 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Starting epoch: SLOW_ADAPTATION, 50 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 3, 3, 2, 1 / 50 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 0, 0, 1, 2 / 50 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 2, 1, 1, 1 / 50 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_04: 1, 1, 1, 0 / 50 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Starting epoch: SLOW_ADAPTATION, 100 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 2, 2, 2, 2 / 100 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 1, 1, 3, 0 / 100 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 1, 2, 1, 1 / 100 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_04: 0, 2, 2, 2 / 100 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Starting epoch: SLOW_ADAPTATION, 200 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 3, 2, 1, 2 / 200 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 1, 0, 2, 2 / 200 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 2, 1, 1, 1 / 200 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_04: 1, 0, 2, 1 / 200 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Starting epoch: SLOW_ADAPTATION, 500 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 2, 1, 3, 4 / 500 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 1, 2, 1, 1 / 500 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 1, 1, 1, 2 / 500 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_04: 1, 1, 3, 1 / 500 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Starting epoch: FAST_ADAPTATION, 50 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 1, 3, 2, 1 / 50 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_01: 1, 1, 0, 0 / 50 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_03: 1, 1, 0, 1 / 50 transitions
liesel.goose.engine - WARNING - Errors per chain for kernel_04: 2, 1, 0, 2 / 50 transitions
liesel.goose.engine - INFO - Finished epoch
liesel.goose.engine - INFO - Finished warmup
liesel.goose.engine - INFO - Starting epoch: POSTERIOR, 1000 transitions, 25 jitted together
liesel.goose.engine - WARNING - Errors per chain for kernel_00: 0, 1, 4, 3 / 1000 transitions
liesel.goose.engine - INFO - Finished epoch

Some tabular summary statistics of the posterior samples:

import liesel.goose as gs

results = engine.get_results()
gs.Summary(results)

Parameter summary:

kernel mean sd q_0.05 q_0.5 q_0.95 sample_size ess_bulk ess_tail rhat
parameter index
concentration_p0_beta (0,) kernel_00 0.069 0.050 -0.011 0.069 0.152 4000 325.513 570.711 1.012
(1,) kernel_00 1.066 0.099 0.905 1.064 1.228 4000 146.567 557.704 1.024
loc_np0_beta (0,) kernel_03 0.600 0.225 0.231 0.599 0.966 4000 46.596 98.020 1.080
(1,) kernel_03 0.323 0.124 0.113 0.324 0.523 4000 99.516 251.131 1.054
(2,) kernel_03 -0.366 0.128 -0.617 -0.354 -0.177 4000 45.433 70.262 1.083
(3,) kernel_03 0.362 0.066 0.248 0.364 0.473 4000 63.030 151.633 1.054
(4,) kernel_03 -0.244 0.086 -0.386 -0.244 -0.098 4000 63.526 156.110 1.089
(5,) kernel_03 0.170 0.031 0.120 0.169 0.222 4000 53.459 132.167 1.097
(6,) kernel_03 6.027 0.037 5.970 6.025 6.091 4000 74.895 249.095 1.032
(7,) kernel_03 0.545 0.072 0.428 0.546 0.662 4000 49.211 178.115 1.106
(8,) kernel_03 1.702 0.029 1.656 1.701 1.751 4000 70.823 211.549 1.034
loc_np0_tau2 () kernel_02 6.286 5.865 2.419 5.091 13.444 4000 3625.394 3754.696 1.000
loc_p0_beta (0,) kernel_04 0.027 0.003 0.023 0.027 0.032 4000 81.008 81.954 1.050
scale_p0_beta (0,) kernel_01 -3.063 0.060 -3.159 -3.064 -2.962 4000 84.917 151.014 1.037
(1,) kernel_01 1.040 0.076 0.914 1.040 1.166 4000 150.826 360.761 1.014

Error summary:

count relative
kernel error_code error_msg phase
kernel_00 90 nan acceptance prob warmup 76 0.019
posterior 8 0.002
kernel_01 90 nan acceptance prob warmup 25 0.006
posterior 0 0.000
kernel_03 90 nan acceptance prob warmup 28 0.007
posterior 0 0.000
kernel_04 90 nan acceptance prob warmup 31 0.008
posterior 0 0.000

And the corresponding trace plots:

fig = gs.plot_trace(results, "loc_p0_beta")


fig = gs.plot_trace(results, "loc_np0_tau2")


fig = gs.plot_trace(results, "loc_np0_beta")


fig = gs.plot_trace(results, "scale_p0_beta")


fig = gs.plot_trace(results, "concentration_p0_beta")

We need to reset the index of the summary data frame before we can transfer it to R.

summary = gs.Summary(results).to_dataframe().reset_index()

After transferring the summary data frame to R, we can process it with packages like dplyr and ggplot2. Here is a visualization of the estimated spline vs. the true function:

library(dplyr)

Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

library(ggplot2)
library(reticulate)

summary <- py$summary

beta <- summary %>%
  filter(variable == "loc_np0_beta") %>%
  group_by(var_index) %>%
  summarize(mean = mean(mean)) %>%
  ungroup()

beta <- beta$mean
X <- py_to_r(model$vars["loc_np0_X"]$value)
estimate <- X %*% beta

true <- sin(2 * pi * x0)

ggplot(data.frame(x0 = x0, estimate = estimate, true = true)) +
  geom_line(aes(x0, estimate), color = palette()[2]) +
  geom_line(aes(x0, true), color = palette()[4]) +
  ggtitle("Estimated spline (red) vs. true function (blue)") +
  ylab("f") +
  theme_minimal()