pygpcca.GPCCA.minChi

GPCCA.minChi(m_min, m_max)[source]

Calculate the minChi indicator (see [Reuter18]) for every \(m \in [m_{min},m_{max}]\).

The minChi indicator can be used to determine an interval \(I \subset [m_{min},m_{max}]\) of good (potentially optimal) numbers of clusters.

Afterwards either one \(m \in I`(with maximal `minChi\)) or the whole interval \(I\) is chosen as input to optimize() for further optimization.

Parameters:

  • m_min (int) – Minimal number of clusters to group into.

  • m_max (int) – Maximal number of clusters to group into.

Return type:

List[float]

Returns:

: List of minChi indicators for cluster numbers \(m \in [m_{min},m_{max}]\), see [Roeblitz13], [Reuter18].