Track down new exception for RBC kalman with fixed x0

[jlperla]

"ME:SingularException: A has eigenvalue(s) with moduli ≈ 1"

Likely just happening because of sampling in divergent regime, but need to check why it is displaying.

[jlperla]

Similarly, getting the following failure when calling with

julia --project --sysimage JuliaSysimage.dll --threads auto bin/fit_rbc_1_kalman.jl --results_path .replication_results/rbc_1_kalman_robustness_0_4_0_4_1_2 --overwrite_results true --num_samples 10 --seed 1 --init_params_file data/rbc_1_kalman_burnin_dynare.csv --override_init_params [0.4,0.4,1.2]

As before, this is in a regime where it shouldn't be working, so this is jsut an error handling issue.

[ Info: Generating 10 samples with 2 adapts across 1 chains using data/rbc_1_kalman_burnin_dynare.csv as initial condition with seed = 1
Sampling 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| Time: 0:00:08
ERROR: LoadError: DomainError with -6.000000000000001:
log will only return a complex result if called with a complex argument. Try log(Complex(x)).
DomainError detected in the user `f` function. This occurs when the domain of a function is violated.
For example, `log(-1.0)` is undefined because `log` of a real number is defined to only output real
numbers, but `log` of a negative number is complex valued and therefore Julia throws a DomainError
by default. Cases to be aware of include:

* `log(x)`, `sqrt(x)`, `cbrt(x)`, etc. where `x<0`
* ```

[jlperla]

Turned back on covariance matrix checks. They seem to help elimiate bad roots well.