7 Conclusion

This work describes the segmentation problem, provides a theoretical model to work with the segmentation of data, and goes on to describe a few algorithms to try to estimate the set of points that optimally segment a given data set, explaining the differences between them.

Then we proceed and describe the actual implementation of such algorithms in an actual R package, in an attempt to make its use easier for researchers. We provide a few applications with simulated data, and also provide a real analysis with Berlin weather data set. Finally, performance considerations are taken when trying to have the code run faster.

In summary, there are a lot of decisions involved in the problem of segmentation, such as picking the right cost function for the evaluation, which algorithm to pick, and whether or not parallelization can help in running the computation faster. By providing a common set of tools, we hope to allow researchers to segment data more easily.