A few of us were chatting at pyastro16 about model errors: it might be useful to add-in the ability to do a bootstrap estimate of model errors. The idea is to do resample the data with replacement, fit the model, and then use the resulting distribution as an estimate of the uncertainty. I made a simple example of the approach here.
A few of us were chatting at pyastro16 about model errors: it might be useful to add-in the ability to do a bootstrap estimate of model errors. The idea is to do resample the data with replacement, fit the model, and then use the resulting distribution as an estimate of the uncertainty. I made a simple example of the approach here.