mefp: Error in root.matrix(crossprod(X)): matrix is not positive semidefinite

I have run into a (rare) case where the root of a matrix cannot be found because it's indefinite. Here's a reprex using the bfast package:

``` r
library(bfast)
BugTS = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 2, 2, 3, 5, NA, NA, 3, 2, 3, 2, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, NA, 0, 1, NA, 0, 1)
BugTS = ts(BugTS, start=2009, frequency = 23)
bfastmonitor(BugTS, start=2016, type = "ME")
#> Error in root.matrix(crossprod(X)): matrix is not positive semidefinite
```

^{Created on 2020-04-07 by the [reprex package](https://reprex.tidyverse.org) (v0.3.0)}

Using @appelmar's C++ code, the same error is rendered as:
```r
Error in .sc_cpp_efp_process_me(X, y, rescale, h) :
sqrtmat_sympd(): transformation failed
```

This fails when `type` is set to `ME`, but it works with any other `mefp` types.

Is there a way to detect that this would happen, and then fall back to some other fluctuation process type instead?