|
|
|
Figure 10
Schematic models for distribution of sampling points. (a)–(c) Three types of model around a point i in high-dimensional space, and (d)–(f) the corresponding hypothesized clustering results obtained by the UMAP analysis, in which dashed rectangles mark the positions of the clusters. The obtained 2D UMAP plots shown in Fig. 6 can be modeled by these distributions. For example, the continuous distribution of clusters 0 to 5 observed in Fig. 6(a) is modeled here by panels (a) and (d), the branches observed in most clusters are modeled here by panels (b) and (e), and the incoherent distribution of the clusters observed in Fig. 6(c) can be modeled here by panels (c) and (f). |
Open
access
access
journal menu![[Figure 10]](nb5399fig10.jpg)



