Clear all

Issues With Homer3 Update

Jordan Borrell
Posts: 10
Topic starter
Active Member
Joined: 2 years ago

Hey Team,

I just updated my Homer3 folder using Sourcetree, so I pulled any updated files from the repository on Github.  After doing this, it appears Homer3 is a bit glitchy overall in general.  Here are a few things of note:

1) Under "Edit Stimulus," there is no longer an option to toggle on/off the stimulus.  Can this option be re-added.  I can see that the amplitude of each stimulus (1, 2, 3, etc.) are either 1 or -1; however, if a stimulus is -1 (toggled off) it still shows a solid vertical line instead of a dashed.  I can change these values manually, but I liked the option of highlighting multiple channels and clicking toggle on/off.  Can this visual (solid vs. dashed stimulus lines) be re-added as well?  It's very helpful visually to see which stimuli are toggled on/off.

2) The GLM no longer works.  Thus, I can no longer analyze my data.  It was working fine until I updated it, as I noticed that there were +33 updates I needed to pull.  I was using the hmr_GLM function, but I now notice that there is a new hmr_GLM_new function.  I've tried both, but both functions give me errors.  Can anyone provide an idea as to why this is now an issue?

I will add to this as I discover new issues. 

Thank you for your help with this!


Topic Tags
3 Replies
Meryem Yücel
Posts: 124
Estimable Member
Joined: 2 years ago

Jordan, could you provide more information about the error you get for the hmr_GLM and hmr_GLM_new functions?




Stephen Tucker
Posts: 52
Trusted Member
Joined: 2 years ago

Hi Jordan,


Thanks for helping us troubleshoot this major update. Please let us know what issues you run into with GLM.


In response to 1)

Stims can still be manually enabled and disabled via a checkbox on the MainGUI panel.

With the checkbox clicked, drawing a rectangle on the data plot will toggle the selected stims.


See these images.


Please note that in previous implementation of SNIRF and more notably the format of .nirs style data structures, the state of a stim as either 0, 1, -1, -2 which specified it as on, off, manually or automatically excluded, etc was referred to interchangeably with a stim's amplitude. This is no longer the case.

Stim states are now stored in the derived data (groupResults.mat) and will not be loaded to or saved from SNIRF files-- the stim's amplitude, which is what you edited in the table, can take any value and is to be used for analysis. See the SNIRF specification for more info.

David Boas
Posts: 217
Estimable Member
Joined: 2 years ago

Jordan, I will add to what Meryem and Stephen said by saying that these were two important changes to Homer3. First, the toggling of stim on and off is now a more formal process. It is not possible to do it in the Homer GUI as was possible in Homer2. We have now add stim data columns that allows you to have much more rich information associated with each stimulus trial. We will soon be perfecting a stim reject or stim inclusion function that will select stim trials based on criteria matches to the stim data columns. It will take a bit of transition to understand this, but in the end will provide much more capabilities.

As for hmrR_GLM_new... big changes made there to include the stim amplitude and stim duration information that is now provided for each stim trial in the stim GUI.

Can you share the specific errors you are getting?