Measuring events with MEDs

I want to do a recording with multiple MEDs during an event that will be an hour or two long. I’m thinking of just taking the number of ones and zeros from each read of a device and graphing the cumulative zscore over the whole period, and doing this separately for each device. And also a combined cumulative zscore similar to how the global consciousness project used to do it.
The question I have is GCP only had 200 bits per sec to deal with. The MEDs obviously have a much higher output and storing all that info, either in memory or on disk, is going to be big, so how to avoid that? I was wondering if there’s any kind of sampling I could do or some other statistical technique I could employ to help avoid this. Perhaps some kind of on the fly calculations to save on memory.

That’s only about 180Mbytes for two generators over two hours. Not a lot in absolute memory size. Your best bet is to store it in RAM and save to disk at the end. Or, save in smaller blocks if RAM is somehow limited. The issue there is you could lose some data while saving to disk unless your program uses two blocks of RAM for each device - one to accept new data and the other to write to disk, alternating back and forth between the two.

You might consider counting the number of 1s in blocks of 8-bit bytes, say 10-100 bytes at a time, but that will limit your flexibility for processing later on. The smaller the block size, the better.

Thank you.
Parking Warden seems to bug out at about 20-30 mins as shown by the sudden downward spike.

I think I’ll work on just recording the raw data and then can post process into graphs and what not later on.
That’ll also save battery on my laptop during the event.

Because you have two separate signals you may want to create an application that integrates the two signals together discriminating their rising and following trending state entrainment. You will save 1/2 the storage and gain a potential coherence metric for mental intention. Further, summing processing is possible on the fly to assess at a slower sampling rate the continuous differential between entrainment states and non-entrainment states, 1s being positive entrainment states and 0’s being negative entrainment states. This differential pattern could be used to drive a right and left application that you would be able to compare active participant mental intention and an ambient environmental state.

One thing that can cause that condition is a memory leak. Eventually the app just runs out of allocated memory, or memory in general. It’s not too hard to see a memory leak, but it can be a little more difficult to fix one.