You are seeing poliastro blog. If you want to see documentation, the latest version can be found in docs.poliastro.space.

So we had to come up with a method which would be computationally efficient and optimal for this kind of problem. That's how LMS, Linear Mean Square, appeared on our door. We used VSSLMS, which is an enhanced LMS version that improves the convergence rate and mean-square error. You might recognize it as one of the main competitors to performance and accuracy of the Kalman Filter, but with the need for less computational power and easier maintenance. What's great about this algorithm is the capability of achieving the solution without the necessity of finding a bracket, since we have no idea which might be the right interval where the answer lies, in this case, we refer to the semi-major axis to acquire a ground-track orbit. Another point of difficulty when trying to model the solution, it's each orbit's particularities. Such as $ecc=0$ or having an orbit with a critical angle or even $inc=0$. Currently, the PR is its last stages of review so let's hope to have this feature soon.