For my application I have been using the batched mode. I have it set to 5. Also is set to a pretty high frequency to get better precision, 15-25 (give or take).
Batch returns a batch of whatever integer you pass in. Example: say you pass in 5, it will return an array of 5 numbers every 5 seconds.
Unless this is a rolling estimation that it makes every 5 seconds wouldn't this be hugely inaccurate?
My logic is coming from the fact that the sensor is reading RR intervals to determine BPM, if you return results after one second the sensor would only catch the High-to-High interval every second if BPM is 60 or less. Any more than this and you just have an estimate... Unless it is taking estimations from the low-to-high points second-to-second and then comparing that to the High-to-High as a baseline once that information comes in. Which may explain inaccuracies at high BPM's.
I am using batched mode to take the batch of points it gives me to average them out to obtain decimal precision. Without decimal precision my app is pretty worthless, I know that default doesn't show decimals at all and is pretty imprecise due to a lower frequency.
Let me know what you think with your experience using both modes!
