tikuma-lsuhsc/python-pvqd

pvqd: Voice Foundation Pathological Voice Quality Database Reader module

Note

This Python package is still under development.

This Python module provides functions to retrieve data and information easily from Voice Foundation's Pathological Voice Quality Database.

This module currently does not retrieve the database itself on its own. User must download and extract the files first from https://prod-dcd-datasets-cache-zipfiles.s3.eu-west-1.amazonaws.com/9dz247gnyb-2.zip

Install

pip install pvqd

Use

from pvqd import PVQD

# to initialize (must call this once in every Python session)
pvqd = PVQD('<path to the root directory of the extracted database>')

# to get a copy of the full database with averaged CAPE-V scores
df = pvqd.query(include_cape_v=True)

# to get age, gender, and mean GRBAS grade scores
df = pvqd.query(["Age", "Gender"], include_grbas='grade')

# to get a dataframe of WAV files and start and ending timestamps of all /a/ segment
df = pvqd.get_files('/a/')

# to iterate over '/a/' acoustic data of female participants along with
# age and mean GRBAS scores
for id, fs, x, auxdata in pvqd.iter_data('/a/',
                                    auxdata_fields=["Age"],
                                    include_grbas=True,
                                    Gender="Female"):
  # run the acoustic data through your analysis function, get measurements
  params = my_analysis_function(fs, x)

  # log the measurements along with the age and GRBAS info
  my_logger.log_outcome(id, *auxdata, *params)