SequentialFeatureSelector: parsimonious model bug?

[trh0ly]

The parsimonious model estimated with SequentialFeatureSelector checks the following condition:

if self.subsets_[k]["avg_score"] >= (max_score - np.std(self.subsets_[k]["cv_scores"]) / self.subsets_[k]["cv_scores"].shape[0]):

but should it no be ...np.sqrt(self.subsets_[k]["cv_scores"].shape[0]))?

See this p. 870 formula 7 for example.

I am happy to be lectured if i am wrong. :)