systole.utils.heart_rate

systole.utils.heart_rate(x: List | ndarray, sfreq: int = 1000, unit: str = 'rr', kind: str = 'cubic', input_type: str = 'peaks') → Tuple[ndarray, ndarray]

Transform peaks data or RR intervals into continuous heart rate time series.

Parameters:

x

Boolean vector of peaks detection or RR intervals.

sfreq

The sampling frequency of the desired output.

unit

The heart rate unit in use. Can be ‘rr’ (R-R intervals, in ms) or ‘bpm’ (beats per minutes). Default is ‘rr’.

kind

The method to use (parameter of scipy.interpolate.interp1d). The possible relevant methods for instantaneous heart rate are ‘cubic’ (defalut), ‘linear’, ‘previous’ and ‘next’.

input_type

The type of input vector. Default is “peaks” (a boolean vector where 1 represents the occurrence of R waves or systolic peaks). Can also be “rr_s” or “rr_ms” for vectors of RR intervals, or interbeat intervals (IBI), expressed in seconds or milliseconds (respectively).

Returns:

heartrate

The heart rate frequency.

time

Time array.

Notes

If the input is in the peaks format, it should be a boolean vector encoding the position of R wave, or systolic peaks.

If it is in the form of RR intervals, it can be expressed in seconds or milliseconds, using rr_s and rr_ms parameters, respectively.

The time and heart rate output will have the same length. Values before the first peak and after the last peak will be filled with NaN values.

Examples

1. From a boolean vector of peaks position:

>>> from systole import import_ppg
>>> ppg = import_ppg().ppg.to_numpy()  # Import PPG recording
>>> _, peaks = ppg_peaks(ppg)  # Find systolic peaks
>>> heartrate, time = heart_rate(peaks)  # Create continuous time series

2. From a vector of RR intervals (miliseconds): >>> from systole import import_rr >>> rr = import_rr().rr.values >>> heartrate, time = heart_rate(rr, unit=”bpm”, input_type=”rr_ms”)