Source number detection

API references

doatools.estimation.source_number.ld_stat(l, n_sources, n_snapshots)[source]

Computes the sufficient statistic for source number detection in MDL/AIC.

Parameters:

  • l (ndarray) – A 1D vector of the eigenvalues of the covariance matrix in ascending order.

  • n_sources (int) – Number of sources.

  • n_snapshots (int) – Number of snapshots.

References

[1] H. L. Van Trees, Optimum array processing. New York: Wiley, 2002.

doatools.estimation.source_number.aic(x, n_snapshots)[source]

Detects source numbers using AIC.

Parameters:

  • x (ndarray) – A 1D vector of the eigenvalues of the covariance matrix in ascending order, or the covariance matrix itself.

  • n_snapshots (int) – Number of snapshots.

Notes

AIC is inconsistent, and tends to asymptotically overestimate the number of sources. However, it tends to give a higher probability of a correct decision.

References

[1] H. L. Van Trees, Optimum array processing. New York: Wiley, 2002.

doatools.estimation.source_number.mdl(x, n_snapshots)[source]

Detects source number using MDL.

Parameters:

  • x (ndarray) – A 1D vector of the eigenvalues of the covariance matrix in ascending order, or the covariance matrix itself.

  • n_snapshots (int) – Number of snapshots.

Notes

MDL is consistent.

References

[1] H. L. Van Trees, Optimum array processing. New York: Wiley, 2002.

doatools.estimation.source_number.sorte(x)[source]

Detects source number using SORTE.

Arg:
x: (~numpy.ndarray): A 1D vector of the eigenvalues of the covariance

matrix in ascending order, or the covariance matrix itself.

References

[1] Z. He, A. Cichocke, S. Xie, and K. Choi, “Detecting the number of clusters in n-way probabilistic clustering,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 32, pp. 2006-2021, Nov. 2010.