Hi Apache Groovy community, I have the pleasure to announce the release of the first beta version of MirComp, the framework for algorithmic and assisted music composition: https://mcolletta.github.io/mircomp/ MirComp is developed in Groovy (currently v4.0.15 with STC by default) and can now run as a self-contained application with OpenJdk 21 and OpenJfx 21. This new version contains lots of bug fixes and improvements but the main highlights of this release are:
MirGene: a Grammatical Evolution (GE) library. One of the most interesting things about this library is its powerful context sensitive grammars that I've designed for musical purposes. A MirGram (as I named it) is stochastic, weighted and adaptable (i.e. it can add new rules during the derivation using the results from previous NonTerminals). Moreover, it can control the length of the results through arithmetic attributes and guards. https://github.com/mcolletta/mircomp/tree/main/mirgene MircOracle: a machine learning library that implements dictionary based predictors and factor oracles (similar to variable order markov models). https://github.com/mcolletta/mircomp/tree/main/mircoracle Note: These two libraries are general enough to be used outside musical contexts. For example in the source of MirGene you can find "symbolic regression" tests. MirSynth: a library for sound synthesis. With few lines of code it is possible to create MIDI "virtual" instruments like Double Frequency Modulation (DFM), Additive and Subtractive Synthesis. Solfeggio syllables and Roman numerals for universal key composition in MirChord. MusicXml Reader: lead sheets music in MusicXml format can now be read and processed by the system. I've tested the software only on Ubuntu (20.04 and 22.04) so if you try the software on other platforms and encounter issues please let me know. Enjoy groovy music and machine learning... in Groovy! Mirco
