# cluster-based correlations in MNE?

**URL:** <https://mne.discourse.group/t/cluster-based-correlations-in-mne/1508>\
**Category:** Mailing List Archive (read-only)\
**Tags:** list-archive\
**Created:** [May 7, 2018, 8:10am UTC](https://mne.discourse.group/t/cluster-based-correlations-in-mne/1508 "2018-05-07T08:10:17Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\
**Post date:** [May 7, 2018, 8:10am UTC](https://mne.discourse.group/t/cluster-based-correlations-in-mne/1508/1 "2018-05-07T08:10:17Z")

</div>

External Email - Use Caution

Hi everyone,

I would like to explore correlations between resting-state EEG frequencies  
power and neuropsychological / behavioral variables and thus look for a way  
to perform cluster-based correlations in MNE, like in this paper  
[https://www.nature.com/articles/srep05101](https://www.nature.com/articles/srep05101)  
\<[https://www.google.com/url?q=https://www.nature.com/articles/srep05101&sa=D&source=hangouts&ust=1525764957803000&usg=AFQjCNFjQb0zLhx\_5sSy3COkbiOH0G728g](https://www.google.com/url?q=https://www.nature.com/articles/srep05101&sa=D&source=hangouts&ust=1525764957803000&usg=AFQjCNFjQb0zLhx_5sSy3COkbiOH0G728g)\>  
(although  
I am open to suggestions for other approaches if you have).

There may be something around the \*permutation\_cluster\_test()\* function but  
it is not clear for me how I have to format my data in this case.

Thank you for your help,

---

<div class="post-metadata">

**Author:** ![system](https://global.discourse-cdn.com/free1/uploads/mne/original/1X/85cc6bd2b69cb698a166dc6d880fb550510d0144.jpeg) [@system](https://mne.discourse.group/u/system)\
**Post date:** [May 9, 2018, 4:50am UTC](https://mne.discourse.group/t/cluster-based-correlations-in-mne/1508/2 "2018-05-09T04:50:05Z")

</div>

External Email - Use Caution

Hi Nicolas,

I am not sure whether a cluster-based correlation test is possible with the MNE-Python stats functions yet, but I implemented it a while ago in Eelbrain which is easy to apply to MNE-Python objects. The relevant function is testnd.corr ([http://eelbrain.readthedocs.io/en/r-0.27/reference.html#module-testnd](http://eelbrain.readthedocs.io/en/r-0.27/reference.html#module-testnd)). Feel free to email me if you want to try it and want more guidance.

Christian