-
-
Notifications
You must be signed in to change notification settings - Fork 390
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Noise simulation #571
Noise simulation #571
Conversation
@zen-juen I'm pretty sure that if we hardcode the SR the results don't change though ^^ could you try pliz and confirm or INFIRM my hypothesis |
@DominiqueMakowski yes yes it doesn't change the actual dimension value (since it's obtained from plotting log power vs freq) but it does alter the frequency axis here in our plot ^^
|
yeah but these values are pretty meaningless anyway, so it's okay (because we didn't specify the unit anyway in the x-axis title so it's still correct). So we can just remove the sampling rate 😁 👍 |
|
@DominiqueMakowski we don't need this ` epochs.append(signal[-window:]) # Add last (smaller) epoch`
@DominiqueMakowski nono because before this line we have
(on hindsight I think this was why I used the |
Interesting, so it's always the first epoch that is eventually a big longer? then it's okay, since it's consistent and only marginally bigger |
korekkorek 😄 |
As for this specific bug in
with
so this will error, though one quick way around it is just to have |
what if we set E to np.nan? would it mess it up downstream? |
What do you mean? Like if there is any d = 0 or dist = 0 then E = np.nan if not E = np.mean(d / dist)? |
yea |
So with your simulated signal example, it'll return
|
I guess it's good as it doesnt seem to break the optimal dimension detection |
No description provided.