enable manually remove bad channels?

[babaq]

Just encounter the situation that some bad channels could be obviously observed when recording, and also when referencing externally, all internal referencing channels are grounded. those channels IDs could be send to KS, so that KS could escape those channels when it start to deciding bad channels, or set them as bad channels anyway after it finished finding bad channels.

I can't find a way to set it right now, and would be trivial to add ops.badch and do it in preprocessDataSub.

any thought? @marius10p

[nsteinme]

Just leave them out of the channel map if you're using the command line (something like: chanMap = chanMap(~ismember(chanMap, badch));). In the GUI you can right-click on the channel in either the data view or the probe view to remove it from analysis.

…

On Thu, Nov 14, 2019 at 5:41 PM Li Alex Zhang ***@***.***> wrote: Just encounter the situation that some bad channels could be obviously observed when recording, and also when referencing externally, all internal referencing channels are grounded. those channels IDs could be send to KS, so that KS could escape those channels when it start to deciding bad channels, or set them as bad channels anyway after it finished finding bad channels. I can't find a way to set it right now, and would be trivial to add ops.badch and do it in preprocessDataSub. any thought? @marius10p <https://github.com/marius10p> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#139>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABZ5IPZZNDTI3NDZLMYJAHTQTX45XANCNFSM4JNUHRIQ> .

[babaq]

yes, i am using matlab functions, and find the ops.chanMap only holding the path to the chanmap files, the loading part and find bad channels are all in the preprocessDataSub function. I could try to do a quick pull request, adding lines you suggested.

[nsteinme]

You can also load the channel map file and pass that in ops.chanMap, rather than the path. Then you can remove the bad channels yourself on the outside. That's the way you should do it.

…

On Thu, Nov 14, 2019 at 6:04 PM Li Alex Zhang ***@***.***> wrote: yes, i am using matlab functions, and find the ops.chanMap only holding the path to the chanmap files, the loading part and find bad channels are all in the preprocessDataSub function. I could try to do a quick pull request, adding lines you suggested. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#139>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABZ5IP7A4USLIXNOOJGYRGTQTX7SDANCNFSM4JNUHRIQ> .

[babaq]

found that loadChanMap could handle loaded chanMap.

Thanks

[babaq]

is it correct to just set the bad channels in ops.chanMap.connected to 0?

[nsteinme]

Yep that should also work!

…

On Thu, Nov 14, 2019 at 6:13 PM Li Alex Zhang ***@***.***> wrote: is it correct to just set the bad channels in ops.chanMap.connected to 0? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#139>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABZ5IP46BXN7LISXQLVDHNDQTYAUFANCNFSM4JNUHRIQ> .

[babaq]

ok, I saw that in preprocessDataSub the total number of channels in chanmap has a chance to overwrite the ops.NchanTOT if it's missing, I am wandering if I removed some bad channels from chanmap, and result in e.g. 269 channels, but the bin file saved all the 384 channels, I should still set ops.NchanTOT=384, or I should leave it unset outside , so it will be set automatically in preprocessDataSub to 269?

[marius10p]

ops.NchanTOT is just the total number of channels in your raw data file.

ops.Nchan gets set by the algorithm after removing bad channels, and is the total number of sorted channels.

[babaq]

Got it, thanks!