PyCuDAL

Parametric acceptance limits for USP <905> Content Uniformity and USP <711> Dissolution.

PyCuDAL is a Python re-implementation of the CuDAL Version 2 SAS application (Bergum, 2007). For a user-specified coverage P (lower bound) and confidence level C, it constructs acceptance-limit tables such that meeting the limits assures any future sample from the batch will pass the applicable USP test at least P % of the time with C % confidence. It also evaluates those tables and computes lower confidence bounds on the probability of passing for observed samples.

Methods

Two sampling plans are supported for each test:

• Plan 1 – one dosage unit per location (e.g. composite QC samples).
• Plan 2 – an equal number of units (> 1) per location (e.g. process validation).

Three analyses are available per plan:

Statistical details are given in docs/USER_MANUAL.md (Appendix B) and in the referenced publications.

Components

Installation

Requires Python ≥ 3.9.

pip install .            # core: numpy, pandas, scipy + CLI + Tk GUI
pip install .[all]       # + matplotlib, reportlab, openpyxl
pip install .[qt]        # + PySide6 (optional Qt GUI)
pip install .[dev]       # + pytest, pyinstaller

Usage

Library

from cudal import cusp1

table = cusp1.acceptance_limit_table(10, 100.0, 95.0, 95.0, 85.1, 114.9, 0.5)
prob  = cusp1.probability_of_passing(table, 10, [95.0, 100.0], [1.0, 4.0])
bound = cusp1.sample_probability(100.0, 2.0, 10, 100.0, 95.0, 95.0)

CLI

cudal                                  # interactive SAS-style menu
cudal cusp1 -m evaluate -o ev.csv      # sub-command mode
cudal disp2 -m sample --mean 90 --se 2.2 --sm 2.46

GUI

cudal-gui                              # Tkinter
python extras/cudal_gui_pyside6.py     # PySide6 (requires pip install .[qt])

See docs/USER_MANUAL.md for a panel-by-panel guide.

Outputs and exports

• Results tables (DataFrames) with sortable columns and conditional formatting.
• CSV per table; XLSX workbook with one sheet per scenario.
• PDF listings reproducing the SAS output format (Courier, wrapped column blocks, SE × SM LL/UL matrices, repeated page headers).
• Plots: points with cubic-spline interpolation, or contour heatmaps with 80 %/90 % threshold contours; exportable as PNG.

Repository layout

cudal/        library: computations, CLI, Tk GUI, package data
extras/       optional PySide6 GUI (excluded from the wheel)
docs/         user manual
tests/        pytest smoke tests
scripts/      executable build helper
.github/      CI, PyPI publishing, release workflows

Testing

pip install -e .[all,dev]
pytest -q
python -m cudal.gui --selftest

References

1. Bergum, J.S. (1990). Constructing Acceptance Limits for Multiple Stage Tests. Drug Development and Industrial Pharmacy, 16(14), 2153–2166.
2. Bergum, J., Utter, M. (2000). Process Validation. In: Encyclopedia of Biopharmaceutical Statistics, Marcel Dekker, 422–439.
3. Bergum, J.S., Utter, M.L. (2003). Statistical Methods for Uniformity and Dissolution Testing. In: Pharmaceutical Process Validation, Marcel Dekker, 667–697.
4. Bergum, J., Li, H. (2007). Acceptance Limits for the New ICH USP 29 Content Uniformity Test. Pharmaceutical Technology, October, 88–98.
5. Bergum, J. (2007). CuDAL Version 2 Users Guide and Appendix E – Lower Bound Calculations.

Acknowledgments

This project is a re-implementation of the CuDAL application originally designed, developed and validated in SAS by Dr. James Bergum. The statistical methods, program structure and reference outputs used here are due to his work; the original SAS programs and validation materials remain the authoritative source.

License

MIT. Provided as-is for research and informational use; it does not constitute regulatory advice. Validate against your own reference implementation before any GMP use.