gnps 0.1.3

Generalized Numerical P Systems simulator package

GNPS project

The GNPS project (Generalized Numerical P Systems) aims to create a library for simulation of different variants of Numerical P Systems (NPS).

It contains:

• GNPS core (the core classes for NPS simulation)
• GNPS parser (parsers for different inputs)
• GNPS generator (generates outputs, i.e. Verilog, Amaranth and Lustre) work in progress.
• Utils (like the main runner)

Input Files

• System description: YAML file describing the GNPS system (required as the first argument).
• Input file: Optional CSV file, each row representing input variable values for a simulation step. The header must match the input variable names defined in the YAML.
• Output file: Optional CSV file, each row representing output variable values for each step.

Command Line Options

Run the simulator with:

python -m gnps <gnps_file.yaml> [input.csv] [output.csv] [options]

Options:

• -c, --compute_mode Run in continuous compute mode (no input variables allowed)

• -s, --steps N Number of steps to run (default: 1)

• --csv Output in CSV format in continuous compute mode

• gnps_file (YAML): Required. GNPS system description.

• input (CSV): Optional. Input values per step (default: stdin).

• output (CSV): Optional. Output values per step (default: stdout).

Modes

• IO mode (default): Reads input variables from CSV for each step, writes output variables for each step.
• Continuous compute mode: Runs for a set number of steps without input variables, outputs results at each step (CSV or text).

Example Input YAML

Below is an example of a GNPS system description in YAML format (from test_math_functions.yaml):

description: "A test system"

cells:
  - id: 1
    contents:
      - x = 0, y = 0, E = 0, z = 0, u = 1
      # Comments can be added like this
    # input variables should be empty for continuous compute mode
    input: [u] 
    output: [x,y,z]

rules:
  - E > x | x + 3 -> x  # guarded rule
  - E + 1 -> E         # unguarded rule
  - 0*z + x + y -> z
  - (E > u || E > x || E > y) | u + x + y -> y                                   # step will be reset to 0

For more details, see the CHANGELOG.md and documentation in the docs/ folder.