Sampling time of intraday Heart Rate Data

[Laxite]

Hi,

 I am pullng intraday heart rate data for an hour with 1sec granularity. However the fastest I have seen so far is new data at 5sec interval. I thought maybe some type of delta compression (output data only if heart rate changes) is employed, but I got data with identical consequitive HR values. So when exactly are the '1sec' granularity values are reported? Is there a way to get it every second? Many HR algorithms calculate 'high-frequency' content till 0.5Hz. For these calculations, the data needs to be available every second or more frequently than that.

Thanks

Jitendra

PS: By the way, this data is from a Charge2

Example data I received is pasted below.

{"value": 88, "time": "19:00:04"}, {"value": 89, "time": "19:00:09"}, {"value": 91, "time": "19:00:19"}, {"value": 90, "time": "19:00:24"}, {"value": 90, "time": "19:00:39"},

[AndrewFitbit]

@Laxite Similar to what I've said here, depends on the data. Although I don't think you can because I haven't seen it consistently return data points for every second. It's usually between 1-3 seconds for me.

[Laxite]

Thanks Andrew. To use this data, there needs to be some predictability. For example, would it be correct to assume that if I did not get the data for a particular second, then it is the same as the last data received and hence it was not sent in the response to save on length of the response? Some knowledge of how the receibed data 'depends on the data' is needed to reconstruct it on the receiving end.

[AndrewFitbit]

@Laxite Yeah I see what you're saying, but don't assume that no data = same data. I agree that there should be some predictability, I'll raise this with the team.

[AndrewFitbit · Accepted solution]

@Laxite Following up on this--

The variability is related to whatever HR data is sent from the device, which can be "up to 1 second". For example, if there's no data between second 5 and second 10, then that means there's no data in between those two points.

If the intervals are close enough, you could average the data between the two points to create another data point.