Error in wfdb_to_wav

Hi everyone,

this is my first time writing an issue as well as using wfdb so, excuse me in advance for the mistakes I'll surely make :)

I'm trying to obtain some .wav from the data stored in VOICED database with the following piece of code:

import wfdb.io.convert.wav
import os

path = 'my_path'
wfdb.io.convert.wav.wfdb_to_wav('voice001', pn_dir='voiced',
                                output_filename=os.path.join(path, 'aaab.wav'))

Thus, I obtain the following error:

Traceback (most recent call last):
File "path_of_my_script", line 18, in
wfdb.io.convert.wav.wfdb_to_wav('voice001', pn_dir='voiced',
File "C:[...]\AppData\Local\Programs\Python\Python310\lib\site-packages\wfdb\io\convert\wav.py", line 121, in wfdb_to_wav
np.left_shift(np.subtract(record.adc(), offset), shift)
File "C:[...]\AppData\Local\Programs\Python\Python310\lib\site-packages\wfdb\io_signal.py", line 578, in adc
nanlocs = np.isnan(self.p_signal)
TypeError: ufunc 'isnan' not supported for the input types, and the inputs could not be safely coerced to any supported types according to the casting rule ''safe''

At first I thought that I was messing it up and went to the documentation page. I ended up realizing that if I try to run the example piece of code that appears for wfdb_to_wav:

wfdb_to_wav('100', pn_dir='pwave')

I obtain the same error as my code's. I did some research and some people has the same 'TypeError: ufunc 'isnan' not supported...' in other context, they managed to fix it using pandas instead of numpy, but it doesn't work for me. Any idea of why is this happening?

Inizia dal punto di ingresso wfdb_to_wav in wfdb/io/convert/wav.py intorno alla riga 121, quindi esamina il metodo adc in wfdb/io/_signal.py intorno alla riga 578. Riproduci l’esempio documentato wfdb_to_wav('100', pn_dir='pwave') e confronta i relativi tipi di input con il percorso che fallisce; il lavoro è completato quando gli esempi vengono convertiti correttamente senza il TypeError segnalato.