v0.1

Initial release

• Create a set of sliders.
• Interactively shape one beta distribution (restricted to a single intuitive "position" parameter rather than independent alpha and beta parameters) for the positive class label and another for the negative class label.
• Interactively specify the class imbalance between the positive and negative classes.
• Dynamically sum the two distributions to get the overall distribution of "model" output scores.
• Dynamically generate a probability calibration curve, given the specified distributions.