Description

Apologies if I'm misreading, but equation 6.7 (p187) seems to suggest the variance of the latent GRW is (1- alpha) * sigma^2 while Code 6.15 (p188) has sigma0 = tf.sqrt(variance * alpha) and seems to use sigma0 as the GRW standard deviation term (z = yield tfd.Sample(tfd.Normal(0., sigma0), num_steps). Apologies if I'm missing something.

Thanks!

Reference
P187/188

Author Section

Do not close until

• Added to Errata
• Fixed in Open Access
• Fixed in latex source