elpigraph.generateInitialConfiguration¶
-
elpigraph.generateInitialConfiguration(X, Nodes, Configuration='Line', DensityRadius=None, MaxPoints=10000, PCADensity=True, CenterDataDensity=True, verbose=False, FixNodesAtPoints=[])[source]¶ Produce an initial graph with a given structure
- X: numerical 2D matrix,
- the n-by-m matrix with the position of n m-dimensional points
- Configuration: string,
type of graph to return. It should be one of the following values: Line:
Points are placed on the 1st principal component between mean-sd and mean+sd- Circle:
- Points are placed on the the plane induced by the 1st and 2nd principal components. In both dimensions they are placed between mean-sd and mean+sd
- Density:
- Two points are selected randomly in the neighborhood of one of the points with the largest number of neighbour points
- DensityProb:
- Two points are selected randomly in the neighborhood of one of a point randomly selected with a probability proportional to the number of neighbour points}
- Random:
- Two points are returned. The first is selected at random, and the second is selected with a probability inversely proportional to thr distance to the 1st point selected
- Nodes: integer,
- number of nodes of the graph
- DensityRadius: numeric,
- the radius used to estimate local density. This need to be set when Configuration is equal to “Density”
- MaxPoints: integer,
- the maximum number of points for which the local density will be estimated. If the number of data points is larger than MaxPoints, a subset of the original points will be sampled
- PCADensity: boolean,
- should PCA be applied to the data before computing the most dense area