More documentation for --tedpca

[benoitberanger]

Summary

Between version 0.0.7 and 0.0.8, the --tedpca parameter changed.
Is it planned to enhance the documentation explaining which tedpca does what ?

Additional Detail

The "decision tree" part in the code is proably the most difficult part

Help in 0.0.7 :

Possible choices: mle, kundu, kundu-stabilize
Method with which to select components in TEDPCA
Default: “mle”

Help in 0.0.8 :

Possible choices: kundu, kundu-stabilize, mdl, aic, kic
Method with which to select components in TEDPCA. PCA decomposition with the mdl, kic and aic options is based on a Moving Average (stationary Gaussian) process and are ordered from most to least aggresive. Default=’mdl’.
Default: “mdl”

Next Steps

Explain a bit more what does each decision tree ?
In particular, why "mdl" is choose to be the default one ?

[tsalo]

@eurunuela You know the PCA methods best at this point. Would you be willing to take this on?

[eurunuela]

Sure, I can tackle this!

[eurunuela]

#551 should take care of this.

[benoitberanger]

Thanks you @eurunuela and @tsalo !

[eurunuela]

I'm going to reopen this as some people still feel like this section of the docs is not very clear.