This example has the potential to be a more informative discussion about the pros/cons of different ICA algorithms. Also right now it doesn't really fit the goal of being a "How-to" for a normal analysis task (you wouldn't usually compare several ICA algorithms every time you preprocess a new dataset). Things it should discuss:
speed
performance on clean data vs different kinds of noisy data
deterministic / random?
would be good to also show the "same" component from different algorithms side-by-side (right now just a separate grid of 25 components for each algorithm)
cmista
commented
3 years ago
This example has the potential to be a more informative discussion about the pros/cons of different ICA algorithms. Also right now it doesn't really fit the goal of being a "How-to" for a normal analysis task (you wouldn't usually compare several ICA algorithms every time you preprocess a new dataset). Things it should discuss:
would be good to also show the "same" component from different algorithms side-by-side (right now just a separate grid of 25 components for each algorithm)