WeakLabelModel

A library for training multiclass classifiers with weak labels

Installation

python3.7 -m venv venv
source venv/bin/activate
pip install --upgrade pip
pip install -r requirements.txt

Run code

See example of a call inside one of the add_queue scripts

Unittest

Run the unit tests running the following script from a terminal

./runtests.sh

Run Jupyter Notebooks

In order to run any notebook in this repository, first create a new kernel that will be available from the Jupyter Notebook.

# Load the virtual environment
source venv/bin/activate
# Create a new kernel
ipython kernel install --name "weaklabels" --user
# Open Jupyter
jupyter notebook

After opening or creating a notebook, you can select the "weaklabels" kernel in kernel -> Change kernel -> weaklabels

Usage

Current usage (may need updating)

Usage: main.py [options]

Options:
  -h, --help            show this help message and exit
  -p DATASETS, --datasets=DATASETS
                        List of datasets or toy examples totest separated by
                        with no spaces.
  -s NS, --n-samples=NS
                        Number of samples if toy dataset.
  -f NF, --n-features=NF
                        Number of features if toy dataset.
  -c N_CLASSES, --n-classes=N_CLASSES
                        Number of classes if toy dataset.
  -m N_SIM, --n-simulations=N_SIM
                        Number of times to run every model.
  -l LOSS, --loss=LOSS  Loss function to minimize between square (brier score)
                        or CE (cross entropy)
  -u PATH_RESULTS, --path-results=PATH_RESULTS
                        Path to save the results
  -r RHO, --rho=RHO     Learning step for the Gradient Descent
  -a ALPHA, --alpha=ALPHA
                        Alpha probability parameter
  -b BETA, --beta=BETA  Beta probability parameter
  -g GAMMA, --gamma=GAMMA
                        Gamma probability parameter
  -i N_IT, --n-iterations=N_IT
                        Number of iterations of Gradient Descent.
  -e MIXING_MATRIX, --mixing-matrix=MIXING_MATRIX
                        Method to generate the mixing matrix M.One of the
                        following: IPL, quasi-IPL, noisy, random_noise,
                        random_weak

Check that all datasets work

The python code in utils/data.py can be run in order to check that all datasets can be downloaded, preprocessed and a classifier can be trained on them. This can be done by calling the python code as a main file

python utils/data.py

Should output the following for each dataset

Testing all datasets
Evaluating iris[61] dataset
----------------
Dataset description:
    Dataset name: iris
    Sample size: 150
    Number of features: 4
    Number of classes: 3
Logistic Regression score = 0.9733333333333334

Upload to PyPi

Create the files to distribute

python3.8 setup.py sdist

Ensure twine is installed

pip install twine

Upload the distribution files

twine upload dist/*

It may require user and password if these are not set in your home directory a file .pypirc

[pypi]
username = __token__
password = pypi-yourtoken

Check the README file for Pypi

In order to check that the readme file is compliant with Pypi standards, install the following Python package

pip install readme-renderer

and run the following command

twine check dist/*

Contributors

• Jesus Cid Sueiro
• Miquel Perello Nieto
• Daniel Bacaicoa