Includes: - 32 avoid NaN for feature_weights_id for new book imports
5.0.0
Includes: - 31 Feature weights set ID per book
4.0.0
Includes: - 29 Treat response textual feature weights as dataset, rather than hard coded defaults that can be overridden. - 30 7.29 Renaming df to datasets, ready for review.
3.1.0
Includes: - 25 add import instructions - 26 Remove not, no, it, and does from the set of pre-filtered stopwords - 27 treat stimulus text as part of question text
3.0.2
This release contains a bugfix to how the `auto` behavior for tag_numeric is reported. Previously, if tag_numeric was auto in the validation request, it was also listed as "auto" in the response. With this change, the boolean return value will indicate if numeric mode was selected.