circ_vtest p value

[idittamir]

Hi, I am trying to use the circ_vtest to check whether the instantaneous phase of local field potential signal (human brain recordings) averages around 180 degrees or not. Since I have 30000 data points for each contact/group, I get either very significant (zero) p value or very non significant (one) values. Because of the high number of samples it seems that even if the mean phase is around 120 and the vector (R) is very short (wide distribution around the mean) - the test still gives low (very significant) p values.

Am I missing something? Is there a better way of doing that? Shouldn't the length of R have a larger impact on the test?

I also tried to use circ_wwtest to compare between 6 different distributions of phase, but I could not find any script for post-hoc multiple comparisons between the grouops assuming the "ANOVA" is significant and the groups are different from one another..?

Thank you! great toolbox!

Idit.

[philippberens]

Did you convert the data to radians?

[idittamir]

yes, it's all in radians (minus pie to plus pie) to begin with, also the dir angle (see the matrix and the script)

Thank you,

Idit

---

From: Philipp Berens notifications@github.com Sent: Tuesday, February 27, 2018 7:50:26 AM To: circstat/circstat-matlab Cc: Tamir, Idit; Author Subject: Re: [circstat/circstat-matlab] circ_vtest p value (#13)

Did you convert the data to radians?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/circstat/circstat-matlab/issues/13#issuecomment-368882698, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Ai00uzatcFZCFVVu3ftHUzI0_X3Btlefks5tZAgigaJpZM4SHW5m.

[philippberens]

Which?

[idittamir]

what do you mean which? the 30000 data points for the instantaneous phase as well as the "dir" angle that I want to test are in radians (pie instead of 180)

[idittamir]

Did you get the script and data I sent you in the previous email?

All data is in radians

---

From: Philipp Berens notifications@github.com Sent: Wednesday, February 28, 2018 1:55:52 AM To: circstat/circstat-matlab Cc: Tamir, Idit; Author Subject: Re: [circstat/circstat-matlab] circ_vtest p value (#13)

Which?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/circstat/circstat-matlab/issues/13#issuecomment-369151599, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Ai00u19kqLWtqGeGuRz-MsVVKdV4kozkks5tZQaIgaJpZM4SHW5m.

[philippberens]

You need to post them on github

On 3/1/2018 6:40 PM, idittamir wrote:

Did you get the script and data I sent you in the previous email?

All data is in radians

---

From: Philipp Berens notifications@github.com Sent: Wednesday, February 28, 2018 1:55:52 AM To: circstat/circstat-matlab Cc: Tamir, Idit; Author Subject: Re: [circstat/circstat-matlab] circ_vtest p value (#13)

Which?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/circstat/circstat-matlab/issues/13#issuecomment-369151599, or mute the threadhttps://github.com/notifications/unsubscribe-auth/Ai00u19kqLWtqGeGuRz-MsVVKdV4kozkks5tZQaIgaJpZM4SHW5m.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/circstat/circstat-matlab/issues/13#issuecomment-369670422, or mute the thread https://github.com/notifications/unsubscribe-auth/ABjr33wqOF3HxkNgkiAsTdiZv5QZIcohks5taDJ-gaJpZM4SHW5m.

-- PD Dr. Philipp Berens (Group Leader) University of Tübingen Institute for Ophthalmic Research

http://www.berenslab.org phone: +49-7071-2988833

[idittamir]

files_for_vtest.zip

[idittamir]

Here are some examples for these files. The test should compare each 30,000 data points (one line at a time) with an angle of pie/minus pie.

Thank you!

[idittamir]

were you able to take a look at the files I uploaded? Thank you!