Python implementation of NARS (Non-Axiomatic-Reasoning-System)

Reference:

• OpenNARS 3.0.4,
• The Design Report of OpenNARS 3.1.0

• Python version: 3.7.10.
  • Only tested under this version, however, Python 3.7 and higher versions maybe acceptable.
• OS: Windows 10.
  • Only tested under this OS, however, other OSs might be ok.
• Packages Requirements: see requirements.txt.
  • It is noted that the version of the python package tqdm should be no higher than 3.1.4, otherwise the color display would be abnormal. This is because of a bug of tqdm, which leads to conflicts between sty and tqdm and cause unexpected color display of sty. However, this constraints is not necessary, i.e., higher version of tqdm is ok if you don't mind abnormal display occuring. The abnormal case only occurs if you first run PyNARS when SparseLUT (Sparse Look-Up Table) is built.

pip install pynars

1. Copy the file pynars/config.json to your workspace-directory. (Optional)
2. In the workspace-directory, run cmd python -m pynars.Console. To execute an *.nal file, run cmd python -m pynars.Console <your-file-name.nal>
3. Input Narsese in the console, input an positive integer to run a number of cycles, or input a comment which is a string with ' as the beginning, e.g. ' your comment.
4. Press ctrl+C to exit.

1. Fork the repository
2. Create Feat_xxx branch
3. Commit your code
4. Create Pull Request

Note: This document will be imporved in the future.