Recently I have started to see a lot of
[1] "Chain 1: Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"
[2] "Chain 1: Exception: cholesky_decompose: Matrix m is not positive definite (in 'chunks/parameters-transformed.stan' at line 20; included from 'model_lgp' at line 40)"
[3] "Chain 1: If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"
[4] "Chain 1: but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."
[5] "Chain 1: "
[6] "Chain 1: Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"
[7] "Chain 1: Exception: cholesky_decompose: Matrix m is not positive definite (in 'chunks/parameters-transformed.stan' at line 20; included from 'model_lgp' at line 40)"
[8] "Chain 1: If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"
[9] "Chain 1: but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."
[10] "Chain 1: "
[11] "Chain 1: Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"
[12] "Chain 1: Exception: Exception: gp_exp_quad_cov: magnitude is 0, but must be > 0! (in 'chunks/functions-kernels.stan' at line 101; included from 'model_lgp' at line 24)"
[13] " (in 'chunks/parameters-transformed.stan' at line 14; included from 'model_lgp' at line 40)"
[14] "Chain 1: If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"
[15] "Chain 1: but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."
[16] "Chain 1: "
[17] "Chain 1: Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"
[18] "Chain 1: Exception: cholesky_decompose: Matrix m is not positive definite (in 'chunks/parameters-transformed.stan' at line 20; included from 'model_lgp' at line 40)"
[19] "Chain 1: If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"
[20] "Chain 1: but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."
[21] "Chain 1: "
[22] "Error in sampler$call_sampler(args_list[[i]]) : "
[23] " Exception: neg_binomial_2_log_rng: Random number that came from gamma distribution is 8.79585e+09, but must be less than 1.07374e+09 (in 'chunks/generated.stan' at line 34; included from 'model_lgp' at line 52)"
error occurred during calling the sampler; sampling not done
with rstan 2.21. Is it common that this happens during warmup and earlier the informational messages were just not printed out? I have seen it also during sampling though. Why can it happen that the magnitude parameter for gp_exp_quad_cov
goes to exactly zero? I have a lower=0
and half-student-t prior for it.