MNE-Python version: 0.23
operating system: Mac OS 11.3.1, Python 3.9
I have a 10 minute recording with sampling frequency 200 hz and have dropped all but the first 9 channels for purposes of visualizing sources after performing ICA. I don’t have any information regarding electrode recording locations.
From the documentation I understand
detect_artifacts is experimental, but I was under the impression that it plots all sources and indicates by highlighting those that can be dropped.
My code so far after dropping all but the first 9 channels from raw:
raw = raw.filter(l_freq=1, h_freq=None) #high pass with 1 Hz cutoff ica = ICA(method='fastica', random_state=180, max_iter="auto") ica.fit(raw)
I wasn’t able to discern which of the 10 sources could be “excluded.”
I instead tried
ica.detect_artifacts(raw,start_find=0, stop_find=180) print(ica.exclude)
The result indicated sources to be excluded were 0, 1, and 6,. This was for the first 180 seconds, and, not surprisingly, by changing
stop_find , I found different values of
ica.exclude. However, I was under the impression that
ica.detect_artifacts should also plot all 10 sources, with the excluded ones not highlighted. Is that correct?
My goal is to look at the graphs produced by
ica.detect_sources over the same time intervals to gain a better sense of where the artifacts are located.