File: AAREADME.txt

Purpose:
  This program runs PCA algorithm using 2D gaussian data.

Files:
  03_pca.py
    - main Python script

Requirements:
  - Python 3.x
  - NumPy

Installation:
  Install required packages using pip:
    python -m venv .venv
    source .venv/bin/activate
    pip install numpy 

How to Run:
  The script supports optional command line arguments.

  1) Run with defaults (1000 samples):

     python 03_pca.py

  2) Specify number of samples:

     python 03_pca.py 5000

Command Line Arguments:
  argv[1] (optional):
    Number of samples to generate (integer).
    If not provided or invalid, DEF_NSAMPLES is used.

Default Settings (in the script):
  DEF_MEAN_0:
    [0.0, 0.0]

  DEF_COV_0:
    [[1.0, 0.0],
     [0.0, 1.0]]

  DEF_MEAN_1:
    [0.0, 0.0]

  DEF_COV_1:
    [[1.0, 0.0],
     [0.0, 1.0]]

  DEF_NSAMPLES:
    1000

  DEF_RANDOM_SEED:
    27
  
  DEF_PCA_NCOMP:
    2
  
  DEF_SYMMETRY_TOL:
   1e-12

Notes:
  1) The covariance matrix must be symmetric. The script checks symmetry using
     np.allclose with DEF_SYMMETRY_TOL.

  2) To change the distribution parameters, edit DEF_MEAN and DEF_COV at the
     top of the script.

Example:
  Generate 2000 samples 
    python 03_pca.py 2000