Skip to content

Latest commit

 

History

History
38 lines (24 loc) · 1.34 KB

README.md

File metadata and controls

38 lines (24 loc) · 1.34 KB

CopyForward

A MIREX project for the 2019 Patterns for Prediction task:

https://www.music-ir.org/mirex/wiki/2019:Patterns_for_Prediction

by Timothy de Reuse.

Input Representations: symbolic, monophonic & symbolic, polyphonic

Subtasks: 1

Command line:

usage: copy_forward.py [-h] [-i INPUT_PATH] [-o OUTPUT_PATH] [-l LIMIT] [-q]

optional arguments:
  -h, --help      show this help message and exit
  -i INPUT_PATH   Path to folder containing .csvs with primes to process.
  -o OUTPUT_PATH  Path to existing folder where predicted continuations will be saved.
  -l LIMIT        Process no more than this number of primes from the input
                  folder (Default: process all files in the input folder).
  -q              Suppress console output.

Example usage:

python copy_forward.py -i ./PPDD-Sep2018_sym_poly_large/prime_csv -o ./copyforward_out -l 100

Each .csv file containing a prime will be processed and saved with the same name in the output folder; the file input/foo.csv will have its predicted continuation saved into output/foo.csv.

Will use only one thread. Memory footprint will be on the order of tens of megabytes (all input .csv files will be loaded into memory simultaneously). Takes about 15 minutes to process 10,000 primes on an ordinary laptop (1.6 GHz Intel i5).

Requires Python 3.7 and numpy >= 1.16.1.