smallerize 0.5.0

A Python implementation of minimisation for clinical trials

smallerize

A Python implementation of minimisation for clinical trials

• Open source: Mozilla Public License 2.0

• Documentation: https://smallerize.readthedocs.io.

• Source: https://gitlab.com/warsquid/smallerize.

Features

• Implements minimization as described in Pocock + Simon (1975): Sequential treatment assignment with balancing for prognostic factors in the controlled clinical trial

• Tested using pytest to ensure the results match the original implementation.

• Pure Python module with no dependencies (pandas is useful when conducting simulations but is optional)

• Includes all functions described in the article: range, standard deviation, variance, etc.

• Also implements the biased-coin minimization method described in Han et al. (2009): Randomization by minimization for unbalanced treatment allocation, to allow for unequal allocation ratios.

• Allows pure random assignment for comparison

• Simulation module to allow simulating the effects of different assignment schemes.

Example

Comparing minimization to purely random assignment by simulation:

See the example notebook for details of the simulation.

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

History

0.5.0 (2019-05-08)

• Improved performance with large numbers of factors/factor levels by making the count table sparse.

0.4.0 (2019-02-02)

• Switched to Mozilla Public License.

0.3.0 (2019-01-02)

• Make it easier to check the valid total imbalance and probability methods.

0.2.0 (2018-10-22)

• Add the biased coin minimization method to allow for unequal treatment allocations.

0.1.0 (2018-10-07)

• First release on PyPI.