Visualizing data using t-SNE

In this script we used t-distributed stochastic neighbor embedding (t-SNE) to visualize animal social niches. We used “Rtsne” package. t-SNE is a dimension reduction algorithm, which transforms high-dimensional data (nine dimensions in our example) to low-dimensional data (two dimensions) which is easier to analyze. This algorithm is an unsupervised dimension reduction algorithm, hence the two axes on the final graph have no units.