Uniform Manifold Approximation and Projection umap : Calculates nonlinear dimension reduction on a data set.
| Input projection | . |
|---|---|
row |
represents the variables (e.g. genes, channels, markers) |
col |
represents the observations (e.g. cells, samples, individuals) |
y-axis |
measurement value |
| Input parameters | . |
|---|---|
init |
character, type of initialization for the coordinates, see details |
scale |
numeric, type of scaling to apply to data |
spread |
numeric, the effective scale of embedded points. In combination with min_dist, this determines how clustered/clumped the embedded points are |
min_dist |
numeric, the effective minimum distance between embedded point |
pca |
numeric, If set to a positive integer value, reduce data to this number of columns using PCA |
prop.train |
numeric, proportion of data used to train the model. The rest of the data will be transformed. Default is 1 (all data is used). |
| Output relations | . |
|---|---|
umap01, umap02 |
first two components containing the new projected values |
The operator performs umap analysis. It reduces the amount of variables (i.e. indicated by rows) to a lower number (default 2). This operators wraps the uwot::umap(). See (https://github.com/jlmelville/uwot) for more details, especially settings and examples.