Skip to content

Latest commit

 

History

History
29 lines (15 loc) · 1.32 KB

README.md

File metadata and controls

29 lines (15 loc) · 1.32 KB

Scripts

Data

prepare_dataset.py: Create a new dataset. See instructions in Create the dataset

create_chroma_vocabulary.ipynb: A notebook that creates a reduced vocabulary of chord types to use for modeling chromas.

MIDI_playback_from_image.ipynb: Piano roll images can be synthesized to audio to using this notebook. It is also available in Colab: Open In Colab.

Analysis

pairspace.py: performs dimensionality reduction over PartPairDatasets using both MDS and t-SNE, and visualizes the resulting low-dimensional spaces

dimensionality_reduction.py: performs dimensionality reduction over PartDatasets using both MDS and t-SNE, and visualizes the resulting low-dimensional space.

analyze_segment_resolutions.py: Answers the question, "What % of segments are at n ticks per beat?"

load_segment.py: Load a single MIDI segment from a dataset.

get_random_segments.py: Select segments from a dataset at random. Just for exploring.

plot_part_pair_dists.py: Create various part pair distribution plots.

Utility

utils.py: Generic helper functions used in multiple scripts

notebook_utils.py: Helper functions for notebooks