Identifies clusters in fast streaming data using a density-based spatial clustering algorithm. Regression planes visualize how each of the 3 dimensions is predicted by the other 2 dimensions.

The camera can orbit the cube and zoom in on it using the mouse, touch or trackpad.

The spheres represent USD markets and their location in a 3D space spanned by 3 data dimensions. Colors encode the membership of spheres to a cluster, with spheres in a cluster sharing the same color, except for gray spheres which do not belong to any cluster. The most important parameter for controlling the size and number of clusters is the `Cluster Radius`

: To generally find more small clusters, decrease it; to generally find fewer large clusters, increase it.

Regression planes visualize how each of the 3 dimensions is predicted by the other 2 dimensions. The axis label of the predicted dimension and the corresponding regression plane share the same color.

The data dimensions of the axes can be changed. `24H Volume USD`

, `24H Trades`

and `Spread %`

are used as the default data dimensions as these are usually the dimensions with the highest absolute correlations, making the usefulness of regression planes very obvious. It is recommended to use `Log`

-transformation for the axis data. If the same data is used for 2 axes, a 2D plane is created; if the same data is used for 3 axes, a 1D line is created.

The spheres representing the USD markets can be hovered over to reveal more information in the top left corner of the screen. When hovering over them, lines appear that indicate the distance from the sphere to the sides of the cube and to the regression planes. The greater the distance from the regression planes, the more of an outlier the market is. The spheres also light up briefly when that particular market receives a data update from the WebSocket API.

To improve performance and save battery power, the cube scene is only rendered on demand, e.g. when the camera is used or new data is received, rather than in a constant animation loop as in most three.js examples.

The material on this website is for information purposes only and does not constitute investment advice.