Skip to main content

The Parkinson`s Disease Data Science Toolkit

Project description



Regular install

$ pip install pdkit


$ pip install git+git://

For “editable” install

$ pip install -e git://

For development install

$ git clone
$ pip install -r requirements.txt
$ pip install .


Example how to use pdkit to calculate tremor amplitude and frequency:

>>> import pdkit
>>> tp = pdkit.TremorProcessor()
>>> ts = pdkit.TremorTimeSeries().load(filename)
>>> amplitude, frequency = tp.amplitude(ts)

where, filename is the data path to load, by default in the cloudUPDRS format.

Pdkit can also read data in the MPower format, just like:

>>> ts = pdkit.TremorTimeSeries().load(filename, 'mpower')

where, filename is the data path to load in MPower format.

To calculate Welch, as a robust alternative to using Fast Fourier Transform, use like:

>>> amplitude, frequency = tp.amplitude(ts, 'welch')

This class also provides a method named extract_features to extract all the features available in Tremor Processor.

>>> tp.extract_features(ts)


>>> import pdkit
>>> ts = pdkit.TremorTimeSeries().load(filename)
>>> tp = pdkit.TremorProcessor(lower_frequency=0.0, upper_frequency=4.0)
>>> amplitude, frequency = tp.bradykinesia(ts)


Example how to use pdkit to calculate various Gait features:

>>> import pdkit
>>> ts = pdkit.GaitTimeSeries().load(filename)
>>> gp = pdkit.GaitProcessor()
>>> freeze_times, freeze_indexes, locomotion_freezes = gp.freeze_of_gait(ts)
>>> frequency_of_peaks = gp.frequency_of_peaks(ts)
>>> speed_of_gait = gp.speed_of_gait(ts)
>>> step_regularity, stride_regularity, walk_symmetry = gp.walk_regularity_symmetry(ts)

where, filename is the data path to load, by default in the CloudUPDRS format.


Example how to use pdkit to calculate the mean alternate distance of the finger tapping tests:

>>> import pdkit
>>> ts = pdkit.FingerTappingTimeSeries().load(filename)
>>> ftp = pdkit.FingerTappingProcessor()
>>> ftp.mean_alnt_target_distance(ts)

kinesia scores (the number of key taps)

>>> ftp.kinesia_scores(ts)


Pdkit can be used to extract all the features for different measurements (i.e. tremor, finger tapping) placed in a single folder. The result is a data frame where the measurements are rows and the columns are the features extracted.

>>> import pdkit
>>> testResultSet = pdkit.TestResultSet(folderpath)
>>> testResultSet.process()

where folderpath is the relative folder with the different measurements. For CloudUPDRS there are measurements in the following folder ./tests/data. The resulting dataframe with all the features processed is saved in testResultSet.features

We can also write the data frame to a output file like:

>>> testResultSet.write_output(dataframe, name)


Pdkit can calculate the UPDRS score for a given testResultSet.

>>> import pdkit
>>> updrs = pdkit.UPDRS(data_frame)

The UPDRS scores can be written to a file. You can pass the name of a filename and the output_format

>>> updrs.write_model(filename='scores', output_format='csv')

To score a new measurement against the trained knn clusters.

>>> updrs.score(measurement)

To read the testResultSet data from a file. See TestResultSet class for more details.

>>> updrs = pdkit.UPDRS(data_frame_file_path=file_path_to_testResultSet_file)

Clinical UPDRS

Pdkit uses the clinical data to calculates classifiers implementing the k-nearest neighbors vote.

>>> import pdkit
>>> clinical_UPDRS = pdkit.Clinical_UPDRS(labels_file_path, data_frame)

where the labels_file_path is the path to the clinical data file, data_frame is the result of the testResultSet.

To score a new measurement against the trained knn clusters.

>>> clinical_UPDRS.predict(measurement)

To read the testResultSet data from a file. See TestResultSet class for more details.

>>> clinical_UPDRS = pdkit.Clinical_UPDRS(labels_file_path, data_frame_file_path=file_path_to_testResultSet_file)

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pdkit-1.3.2.tar.gz (47.9 kB view hashes)

Uploaded source

Built Distribution

pdkit-1.3.2-py2.py3-none-any.whl (58.3 kB view hashes)

Uploaded py2 py3

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page