Lines indicate distance from true power profile at a certain number of clusters used in lasing off profile. Dots indicate the number of clusters chosen by clustering algorithm.
Cosine and l1 are both hierarchical clustering methods with cosine and l1 distance metrics (respectively). Kmeans uses the kmeans algorithm. Hierarchical is hierarchical clustering with a Euclidean distance metric. North coast uses the upper 'coast' in image to cluster images. 'Old' uses the old grouping method.
Same as above but includes baseline for perspective. Baseline is generated by choosing random lasing off profile from lasing off run to estimate power.
Overview
Content Tools