* Fixed issue with the computation of the KL divergence
2.1.3
* Library tested on Python 3.7 => removed requirements for Python 3.9 only. Now accepts Python >=3.7
2.1.2
* Add a function to help when building the DAG.
2.1.1
* Add a way to enforce the bounds instead of discarding values outside of bounds. (Useful for mixed distributions with peak of values close to one of the bounds.)
2.1.0
* Add two types of conditionality to the DATGAN (rejection sampling and conditional inputs)
2.0.1
* Add a new way to define the data types to better control continuous columns