Structure of tsam

Tsam consists of a main class containing the basic functionailities to aggregate time series to typical periods based on averaging, k-means clustering of hierarchical clustering. The utils-folder contains additional clustering algorithms including a mixed-interger linear programming based formulation of the k-medoids algorithm and constrained agglomerative clustering of adjacent time steps called segmentation.

Submodule with the basic functionalities

• Time Series Aggregation Class
• Hypertuning segment and period selection
• Aggregation function
• Representation function

Sub functions such as k-medoids and segmentation

• Exact k-Medoids Functionality
• k-Maxoids Functionality
• Duration Representation Functionality
• Segmentation Functionality