Input-output

pyPCG.io.read_signal_file(path: str, format: str) → tuple[ndarray[tuple[int, ...], dtype[int64]], int]

Read in fetal heartsound containing file

Supported file formats:

• wav: wav file

• mat: MATLAB file (containing two variables: fs-samplerate, sig-signal data)

• raw: raw binary (headerless)

• FETA: every second byte is PCG data (headerless)

• 1k: 1 kB chunks (CURRENTLY UNTESTED)

Parameters:

• path (str) – Path to input file

• format (str) – File format identification

Returns:

Unprocessed heartsound signal read in from file, and sample rate in Hz. If the input file was headerless, then the value is 0

Return type:

tuple[np.ndarray,int]

Example

Read in a 1 min wav file with 333 Hz samplerate

>>> import pyPCG.io as pcg_io
>>> data, fs = pcg_io.read_signal_file("example.wav","wav")
>>> print(len(data),fs)
19980 333

Raw binary file reading: >>> import pyPCG.io as pcg_io >>> data, fs = pcg_io.read_signal_file(“example.dat”,”raw”) >>> print(len(data),fs) 19980 0

pyPCG.io.read_hsannot_file(fpath: str) → tuple[list[float], list[float]]

Reads manually labeled heartsounds from annotation csv file

The csv format has to be the following:

• Location: heartsound location in seconds

• Value: heartsound type of “S1” or “S2”

Parameters:

fpath (str) – path to annotation file

Returns:

S1 and S2 annotation locations

Return type:

tuple[list[float],list[float]]