Version 1.0.0 is the first public release of the code accompanying our manuscript on the Density-Matrix Method (DMM) for high-cardinality categorical learning. This release provides a fully reproducible implementation of the proposed pipeline together with the synthetic data generators used to evaluate controlled regimes. Main features • End-to-end Jupyter notebook (dmm_synthetic_experiments.ipynb) implementing: • categorical block-based synthetic generators, • density-operator construction from class-conditional frequencies, • spectral embedding and low-rank surrogate representations, • class-conditional KDE classification in the latent space, • baseline comparisons (PCA+KNN and Random Forest). • Synthetic experiments S1–S4: • S1: controlled class separation via overlap parameter \delta, • S2: increasing cardinality/sparsity by scaling modality sizes, • S3: robustness to irrelevant categorical blocks (noise ratio \alpha), • S4: DMM-only ablation under class imbalance (ML vs MAP decision rules). • Reproducible execution via a conda environment specification (environment.yml). Notes All datasets are fully synthetic and generated procedurally from fixed random seeds. This release is intended to support transparency and reproducibility of the experimental section of the manuscript.