plotting dendogram

brief description: create a plot showing groups/clusters of architects who were born around similar times (i.e. a range of year)

method: there are variety of ways to create a plot that show group and cluster relationships

• dendogram
• parallel coordinate
• cluster plot
• orthogonal convex hull (a potential method)

in this post, dendogram is explored

using scipy library:

learning:

since I started this plotting exercise, I have been looking for a solution of plotting all data I have gathered into plots that show interconnected group and cluster relationships. dendogram is a way of showing that. these initial results are definitely intriguing. the plots are the first time I am able to combine separate category of data – architects by DOB and science by pub.date.

there are still a few concepts I need to learn to gain a full understanding of how dendogram function in scipy works. the next step would be to learn more about the algorithm behind how the clustering is defined. the algorithm defines how accurately, closely the clusters are sorted. using scipy dendogram function, the ‘how’ is not apparent.

another technique i would like to look into is ‘orthogonal convex hull’. this could be a way of plotting cluster/groups in a more spatial way.

references:
https://scipy.github.io/devdocs/generated/scipy.cluster.hierarchy.dendrogram.html
https://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.set_link_color_palette.html#scipy.cluster.hierarchy.set_link_color_palette
https://stackoverflow.com/questions/50846621/how-to-adjust-branch-lengths-of-dendrogram-in-matplotlib-like-in-astrodendro
https://bradleyboehmke.github.io/HOML/hierarchical.html
https://towardsdatascience.com/https-towardsdatascience-com-hierarchical-clustering-6f3c98c9d0ca
https://towardsdatascience.com/machine-learning-algorithms-part-12-hierarchical-agglomerative-clustering-example-in-python-1e18e0075019
https://docs.scipy.org/doc/scipy/reference/generated/scipy.cluster.hierarchy.linkage.html