reverberation_time_linear_regression divide by zero warning when last time sample of IR is 0

[tjueterb]

Is there an existing issue for this?

• I have searched the existing issues

Problem description

Hi all,

just a small inconvenience: If the last sample of an impulse response is 0 (e.g. because it's windowed), the last sample of the EDC (from schroeder_integration is also 0 and reverberation_time_linear_regression throws a divide by zero warning. I think it wouldn't be too harmful to supress that warning.

Code sample

Code run:

ir = pf.signals.files.room_impulse_response()
# fade out of the last two samples
ir = pf.dsp.time_window(ir, interval=(ir.n_samples-2, ir.n_samples-1),shape='right')
edc = pr.edc.schroeder_integration(ir, is_energy=False)
t30 = pr.parameters.reverberation_time_linear_regression(edc, T='T30')
print(f'T30: {t30} seconds')

Traceback:

T30: [2.00420347] seconds
/home/tjueterb/code/misuka/.venv/lib/python3.12/site-packages/pyrato/parameters.py:83: RuntimeWarning: divide by zero encountered in log10
  edcs_db = 10*np.log10(np.abs(energy_decay_curve.time))

Environment

Details

I pulled the develop branch just before testing. ```text pyrato==0.4.1 ├── pyfar [required: >=0.5.0, installed: 0.7.4] │ ├── numpy [required: >=1.23.0, installed: 2.4.2] │ ├── scipy [required: >=1.5.0, installed: 1.16.3] │ │ └── numpy [required: >=1.25.2,<2.6, installed: 2.4.2] │ ├── matplotlib [required: Any, installed: 3.10.8] │ │ ├── contourpy [required: >=1.0.1, installed: 1.3.3] │ │ │ └── numpy [required: >=1.25, installed: 2.4.2] │ │ ├── cycler [required: >=0.10, installed: 0.12.1] │ │ ├── fonttools [required: >=4.22.0, installed: 4.61.1] │ │ ├── kiwisolver [required: >=1.3.1, installed: 1.4.9] │ │ ├── numpy [required: >=1.23, installed: 2.4.2] │ │ ├── packaging [required: >=20.0, installed: 26.0] │ │ ├── pillow [required: >=8, installed: 12.1.1] │ │ ├── pyparsing [required: >=3, installed: 3.3.2] │ │ └── python-dateutil [required: >=2.7, installed: 2.9.0.post0] │ │ └── six [required: >=1.5, installed: 1.17.0] │ ├── sofar [required: >=0.1.2, installed: 1.2.3] │ │ ├── netCDF4 [required: !=1.7.4, installed: 1.7.3] │ │ │ ├── cftime [required: Any, installed: 1.6.5] │ │ │ │ └── numpy [required: >=1.21.2, installed: 2.4.2] │ │ │ ├── certifi [required: Any, installed: 2026.1.4] │ │ │ └── numpy [required: Any, installed: 2.4.2] │ │ ├── numpy [required: >=1.14.0, installed: 2.4.2] │ │ ├── beautifulsoup4 [required: Any, installed: 4.14.3] │ │ │ ├── soupsieve [required: >=1.6.1, installed: 2.8.3] │ │ │ └── typing_extensions [required: >=4.0.0, installed: 4.15.0] │ │ ├── requests [required: Any, installed: 2.32.5] │ │ │ ├── charset-normalizer [required: >=2,<4, installed: 3.4.4] │ │ │ ├── idna [required: >=2.5,<4, installed: 3.11] │ │ │ ├── urllib3 [required: >=1.21.1,<3, installed: 2.6.3] │ │ │ └── certifi [required: >=2017.4.17, installed: 2026.1.4] │ │ └── packaging [required: Any, installed: 26.0] │ ├── urllib3 [required: Any, installed: 2.6.3] │ ├── deepdiff [required: Any, installed: 8.6.1] │ │ └── orderly-set [required: >=5.4.1,<6, installed: 5.5.0] │ └── soundfile [required: >=0.11.0, installed: 0.13.1] │ ├── cffi [required: >=1.0, installed: 2.0.0] │ │ └── pycparser [required: Any, installed: 3.0] │ └── numpy [required: Any, installed: 2.4.2] ├── numpy [required: >=1.14.0, installed: 2.4.2] ├── scipy [required: >=1.5.0, installed: 1.16.3] │ └── numpy [required: >=1.25.2,<2.6, installed: 2.4.2] └── matplotlib [required: Any, installed: 3.10.8] ├── contourpy [required: >=1.0.1, installed: 1.3.3] │ └── numpy [required: >=1.25, installed: 2.4.2] ├── cycler [required: >=0.10, installed: 0.12.1] ├── fonttools [required: >=4.22.0, installed: 4.61.1] ├── kiwisolver [required: >=1.3.1, installed: 1.4.9] ├── numpy [required: >=1.23, installed: 2.4.2] ├── packaging [required: >=20.0, installed: 26.0] ├── pillow [required: >=8, installed: 12.1.1] ├── pyparsing [required: >=3, installed: 3.3.2] └── python-dateutil [required: >=2.7, installed: 2.9.0.post0] └── six [required: >=1.5, installed: 1.17.0]

</details>


### Anything else?

_No response_