CSV data cleaning

csv

#1

Having recently gotten muse player up and running, I have transformed my muse recording into CSV. As some will be aware the data in CSV format is neither very clean nor very intuitive.

It seems to me the first steps could be to manually split the lines into columns and then fix the time scale afterwards. There are some threads and some documentation going through the order of the data etc.

My question here is: is there some easy/standard way to clean this data? It seems quite bothersome to split fields and lookup in the manuals one value at a time - so would be nice to avoid that.


#2

If you record with Muse Monitor, I use a much easier to graph CSV format with each data item in it’s own column.

Sample CSV data is available here. This was recorded with 1 second intervals, but in the app you can set it to constant (256Hz) if required.

You can graph the attached sample on the Muse Monitor website, or using the Excel macro here.


#3

Thanks for the answer. That does indeed look better.

I would prefer to go at it without buying an app though.


#4

What I’ve done is import the .CSV file into Excel to get the data into columns. I then filter the column with all the names of the values (/muse/eeg/, /muse/elements/alpharelative, etc…) to isolate what I want to look at. If I want to graph something I’ll copy and paste the filtered data into a seperate worksheet. Usually what’s easiest to look at is the various relative power bands (alpharelative, betarelative, gammarelative, etc…)

As far as fixing the time scale you’d have to look more into the documentation as I’ve not done that with Muse Player .CSV recordings, I’ve been using apps written for NeuroTechX for ERP stuff.

https://github.com/NeuroTechX

It can be a hassle working with .CSV recordings so Muse Monitor app might make things easier.