v1.0.1

• fixed a bug where forests could not be created with unbounded depth
• fixed a bug where doc strings were not correctly rendered in external docs
• fixed a bug where fitting new data to a forest would add new trees to the old fitted forest instead of overwriting them
• fixed a bug where the predict(forest, ...) function would only allow specific types in arrays
• deleted LLM prompts folder, you can still find it in the release before this one

Final

We proudly present our final releases of the OneTwoTree package.
We have achieved the following progress:

• DecisionTrees: Classification and Regression with numerical and categorical input data
• Cart Algorithm for DecisionTree Construction
• The implementation is largely independent of the chosen splitting criterion.
• Gini impurity
• Information gain
• Variance splitting criteria
• Custom splitting criteria as arguments possible
• Random Forests: Classification and Regression with numerical and categorical input data
• Improved performance,
• Nice documentation
• Improved ReadMe,
• Better demo examples
• Dependency-free
• 1 0 0 % code coverage

Milestone 2 Release

• DecisionTrees: Classification and Regression with numerical and categorical input data
• Cart Algorithm for DecisionTree Construction
• Input data is now not limited to String/Float anymore
• Random Forests
• Gini impurity, information gain and variance splitting criteria
• Can define own splitting criteria and give it to the fit! function of DecisionTrees (wip to make it more accessible)
• Fixed issues from milestone 1 code reviews
• Improved demo examples for forests/trees
• Improved ReadMe
• maintained > 80% code coverage through tests

Milestone 1 Release

• DecisionTrees: Classification and Regression with numerical input data
• Cart Algorithm for DecisionTree Construction
• Gini impurity and information gain
• ReadMe instructions
• Utility functionality such as saving and loading datasets to and from csv
• > 80% code coverage through tests