v1.0.0

[2019/08/14]

CausalLift version 1.0.0 adopted Kedro to add the following new
features.

• [Parallel execution] Train the 2 models in parallel
• [File management] Save and load intermediate files such as the trained models
• [Documentation] Generate the API document by Sphinx and visualize the process flow

Other enhancements include:

• [Logging] Show and/or log processing status such as timestamp and the running task
• [Model options] Specify models other than XGBoost and Logistic Regression for uplift
  modeling and propensity modeling, respectively.

v0.0.3

[2019/04/29]
Merged pull request #2, thanks @farismosman !

• Add unit testing utility functions in the module named utils.
• Include a minor bug fix for the function overall_uplift_gain_ where col_treatment and col_outcome might not default to 'Treatment' and 'Outcome' respectively.
• Import missing numpy module.

v0.0.2

[2019/04/29]

• Add simple test codes.
• Add "from IPython.display import display" so it can run in non-IPython environments.
• Fix "TypeError: object of type 'float' has no len()" that occurs if enable_ipw is set to False (#1, thanks @farismosman !)
• [Examples] Skip installation of CausalLift if not run in Google Colab.

v0.0.1

[2019/04/04]
Initial release