Average Band Powers
Tagged: VALUES MARGIN
April 10, 2017 at 6:21 am #276919
I am running the average band powers example in python on ubuntu. Emotiv EPOC+
I was wondering what are expected values and what the units are of this example?
I am getting this sort of output from the console
– 307337.383470, 792445.055177, 550000.162273, 424424.400943, 504742.084554
– 387174.849413, 291974.736747, 395646.891139, 671188.730325, 649236.926420
Could you please advise what the best way is to make sense of this data?
April 21, 2017 at 12:36 am #277545
those values you are getting are much too high.. Sound like artefacts or perhaps some sort of signal problem. normal values from the average band powers example should be much smaller (between 0.0 and 20.0). Very strong brain activity can lead to higher values, but certainly not that high…
Are you getting those values all the time or only during the first few seconds?
April 21, 2017 at 12:38 am #277546
Thanks for your reply! I am getting these values constantly. I am using the python implementations of the average band powers code.
April 21, 2017 at 12:42 am #277547
do you have a good signal quality with all sensors of your EPOC+? I’m not an EMOTIV expert, but I presume it has something to do with the signal or contact quality. Perhaps you can try the java example, that one works fine for me, haven’t tried the python example yet…
April 21, 2017 at 12:48 am #277549
I shall try that now, I check on the app every time I start data capture and the signal quality is at 80-100 percent.
I shall try the java example and revert asap
April 21, 2017 at 12:52 am #277550
good luck with it. I use Eclipse for the java examples. Don’t forget to do the necessary preparations beforehand. You can find them in the community-sdk-folder >> /doc/How to run Java examples.pdf
April 21, 2017 at 6:32 am #277570
Just tried the java examples and it worked fine! I’m getting values such as 0.05713714945137591, 0.043597699801918216, 0.05413118330108578, 0.0543804819466459, 0.02522513543101461, which seem quite reasonable.
I am just wondering why the Python example does not work! I really need to use the python examples as the rest of the software is in python 🙁
Any chance you could have a peek to see if its something obvious wrong in the python example?
April 21, 2017 at 7:59 am #277574
April 21, 2017 at 8:33 am #277579
I am using the latest code, however I have realized something just now!
Testing on Windows 8 64bit – The AverageBandPower java example outputs sane numbers – 0.24663784116116627, 2.4934585148536206, 0.08210989087394555, 0.0777969184574668, 0.05834042449774537,
Testing on Linux Ubuntu – Python 2.7 The example outputs -– 1120.716529, 1240.327180, 1381.966493……
Testing on Linux Java – The example also outputs – the same range of values as python 1120.716529, 1240.327180, 1381.966493.
My question is why do the libraries on different platforms output different values. The windows and linux tests were conducted back to back without moving the emotiv. Something wrong with the linux libraries?
April 24, 2017 at 3:00 am #277733
I’m afraid I can’t answer that question… Looks like something is wrong with the Linux libraries, but can’t tell for sure. Perhaps it’s only a point shifting issue and the values are correct if you move the point. Perhaps write an email to hello AT emotiv DOT com or raise a ticket on their website, maybe the technical support will be able to answer that question.
September 7, 2017 at 6:36 am #286306
Do you mind if I ask how this issue evolved for you? I am having the same issues with the python code on Linux, and am unsure how far those out of range values can be trusted.
September 7, 2017 at 9:16 am #286319
Bad news for you, I have never managed to get any meaningful data from the device using the SDK. Also official support tickets took 4 months to return to me with no useful information.
It seems they are quite happy to sell you the device and offer little to no support for it.
September 13, 2017 at 2:39 pm #286741
Sorry to hear that. On my side, I can confirm that I was able to run the AverageBandPowers example in Java on Windows and it seems to return reasonable values. Whether they’ll be useful remains to be seen. Getting some of the other example scripts to run in all likelihood will be an uphill battle. But I can confirm this one thing worked, and that Windows looks to be the way to go with this device.
September 17, 2017 at 2:52 am #286943
De la Morena AlexiaParticipant
Hi guys, I read your post…and I still wandering the meaning of those values:
Could you help me to interpreted 😉 please?
I mean what are margin values…(from X to X), values of significance…here the participant is doing a bling beverage testing.
COUNTER INTERPOLATED AF3 F7 F3 FC5 T7 P7 O1 O2 P8 T8 FC6 F4 F8 AF4
23 0 4155,384277 4150,256348 4157,94873 4159,487305 4154,358887 4157,94873 4158,461426 4165,12793 4170,769043 4101,025391 4151,794922 4070,256348 4168,717773 4166,153809
24 0 4166,666504 4171,281738 4154,871582 4157,436035 4171,794922 4153,846191 4155,897461 4163,589844 4167,179199 4134,871582 4171,281738 4081,53833 4160 4163,07666
25 0 4149,743652 4135,897461 4161,538574 4161,538574 4134,358887 4157,94873 4157,94873 4148,717773 4149,230469 4073,333252 4121,538574 4052,820313 4149,743652 4156,410156
You must be logged in to reply to this topic.