search

grero / hmmsort.py

Various methods to facilitate spike sorting using an HMM algorithm
GNU General Public License v3.0
1 stars 13 forks source link

Check that the superposition waveform exceed the noise threshold #25

Open shihchengyen opened 5 years ago

grero commented 5 years ago

I have opened some pull-requests for the 3 issues we discussed today. As soon as I have a sample of a data where these problems are apparent, I can check whether these fixes are sufficient.

shihchengyen commented 5 years ago

Here is some data (vmhighpass.mat & hmmsort.mat):

https://cortex.nus.edu.sg:8399/sharing/l4Bao1FH3 https://cortex.nus.edu.sg:8399/sharing/qfzO4jv1j

You can view the highpass with spikes by doing:

load vmhighpass.mat InspectGUI(vh2,'LoadSort')

You will have to clone the Hippocampus repository to get the code for vmhighpass.

shihchengyen commented 5 years ago

Hi Roger,

That was fast!

I have shared some data with you in the GitHub comments.

Let me know if you have any problems working with the data.

Thanks!

Shih Cheng

On Dec 20, 2018, at 5:48 PM, Roger Herikstad notifications@github.com<mailto:notifications@github.com> wrote:

I have opened some pull-requests for the 3 issues we discussed today. As soon as I have a sample of a data where these problems are apparent, I can check whether these fixes are sufficient.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/grero/hmmsort.py/issues/25#issuecomment-448937388, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AdfZh8RJdlIaQnr4Mnj6K2ie6a_NzoG4ks5u61zhgaJpZM4Zbks1.

Important: This email is confidential and may be privileged. If you are not the intended recipient, please delete it and notify us immediately; you should not copy or use it for any purpose, nor disclose its contents to any other person. Thank you.

grero commented 5 years ago

Hm, do you usually save the data in pre v7.3 format? I try to always use -v7.3 when saving so that everything is saved in an HDF5 compatible format. It makes it easier to read in python. I'm actually a bit surprised that sorting this file worked in the first place, since I have trouble loading it now..

shihchengyen commented 5 years ago

Roger,

I think I encountered some issue that was fixed with the older format, but I can check to see if we still need it.

I also changed the data to single precision floats, so I wonder if that is causing problems in Python?

Shih Cheng

On Dec 21, 2018, at 9:32 AM, Roger Herikstad notifications@github.com<mailto:notifications@github.com> wrote:

Hm, do you usually save the data in pre v7.3 format? I try to always use -v7.3 when saving so that everything is saved in an HDF5 compatible format. It makes it easier to read in python. I'm actually a bit surprised that sorting this file worked in the first place, since I have trouble loading it now..

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/grero/hmmsort.py/issues/25#issuecomment-449204991, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AdfZh11sf1t5j5hnRspinZwd0MWmtdwbks5u7DoIgaJpZM4Zbks1.

Important: This email is confidential and may be privileged. If you are not the intended recipient, please delete it and notify us immediately; you should not copy or use it for any purpose, nor disclose its contents to any other person. Thank you.

grero commented 5 years ago

OK, so it turns out that using single instead of double confuses matlab. hmm_decode returns a log likelihood of NaN when the data has single precision.