We released a new Wave_clus version that uses a variable number of features (based on the degree of non-gaussianity of their distribution) and select automatically clusters at different temperatures.
Changes in the Temperature Plot
A cross symbol (‘x’) will be shown on the clusters already analysed, any cluster analysed and not selected have some of their spikes in one of the selected cluster at higher temperature. A grey zone it can be present as well, identifying the temperatures outside the super-paramagnetic regime according the algorithm, usually a manual selection of clusters in that zone will generate overclustering.
We released a new Wave_clus version that uses a variable number of features (based on the degree of non-gaussianity of their distribution) and select automatically clusters at different temperatures.
Changes in the Temperature Plot
A cross symbol (‘x’) will be shown on the clusters already analysed, any cluster analysed and not selected have some of their spikes in one of the selected cluster at higher temperature. A grey zone it can be present as well, identifying the temperatures outside the super-paramagnetic regime according the algorithm, usually a manual selection of clusters in that zone will generate overclustering.
Reference: https://www.physiology.org/doi/full/10.1152/jn.00339.2018