Lab6: clarifications

[jrasero]

• Add a mention to how data should be format for this type of analysis (as a structured array). Also mention that scikit-survival also provides a class to do this from numpy arrays or dataframes directly. (https://scikit-survival.readthedocs.io/en/stable/api/generated/sksurv.util.Surv.html)

• The model without penalization fails because of multicollinearity issues. That is why with penalization works. For the future, either mention this is an expected behavior, but they need to understand why. Or provide a cleaner version of the data without collinearity.