Aplos 1.1

A Linear Problem parser

Aplos is a simple and elegant linear problem(LP) parser. It allows one to get all the information they need about any linear problem given with the correct syntax. You can read more about Linear Programming here.

Expected LP format

min/max c^Tx

s.t /st /s.t. /subject to Ax ⊗ b

End

Where ⊗ can be any of the following =, <=, >=

Variable(x) constraints/domains are not taken into consideration (not yet).

---

Examples:

1. Max 3x1 +2x2

   s.t. x1+2x2<=9

   2x1+5x2<=4

   End

2. min 3x1 - 5x2 + x4
   st x2 + x3 = 2
   2x1 + 3x2 + 5x4 >= 5
   x1 - 5x2 + 2x3 - 4x4 <= 10
   END

Usage

import Aplos

# Initialization
# From a file
parser = Aplos.AplosParser(filename='lp.txt')

# From a string
text_lp = '''Max 3x1 +2x2

s.t. x1+2x2<=9

2x1+5x2<=4

End'''

parser = Aplos.AplosParser(text=text_lp)

# From a string with a custom delimeter
text = "Max 3x1 +2x3 + x5,s.t. x1+2x2<=9,2x1+5x2<=4,End"

parser = Aplos.AplosParser(text=text, delimeter=',')


# Getting the variables
variables_of_line = parser.get_vars(line_idx=0)
# variables_of_line = {"existing":['x1','x3'], "extended":['x1','x2','x3','x4','x5']}
variables_all = parser.get_vars()
# variables_all = ['x1','x2','x3','x4','x5']

# Detect errors
errors = parser.detect_errors() # set print_msg=True to print the full list of errors

if not errors:
    # Get dimensions
    dimensions = parser.get_dimensions()
    m = dimensions['m']
    n = dimensions['n']

    # Get any matrix (A,b,c,Eqin or MinMax)
    # Eqin and MinMax have values corresponding to symbols
    # Eqin -- '<=': -1 | '>=' : 1 | '=' : 0
    # MinMax -- 'max':1 | 'min':-1
    matrix_A = parser.get_matrix('a')
    matrix_b = parser.get_matrix('B')
    # And so on

    # Otherwise, get all matrices at once.
    # Keys are : A,b,c,Eqin & MinMax
    matrices = parser.get_matrices()
    matrix_A = matrices['A']
    # And so on

    # Save matrices to file
    parser.write_matrices_to_file('output.txt')

    # Get dual matrices
    # Variable constraints -- 'free' : 0 | '>= 0' : 1 | '<= 0' : -1}
    dual_A = parser.get_dual_matrix('a')

    # Variable constraints are calculated assuming that x(i) >= 0
    # for every i. This is subject to change.
    dual_var_constr = parser.get_dual_matrix('var_constr')
    # And so on

    # You can also get all the dual matrices together
    # Similarly keys are : A,b,c,Eqin,MinMax & VarConstr
    dual_matrices = parser.get_dual_matrices()
    dual_A = dual_matrices['A']
    # And so on

    # Save dual matrices to file
    parser.write_matrices_to_file('output_dual.txt', dual=True)


    # After saving matrices (non-dual), you can also read them back
    saved_matrices = parser.read_matrices_from_file('output.txt')

    # If dual
    saved_d_matrices = parser.read_matrices_from_file('output_dual.txt')

As the project continues, the 'usage' section will get updated and eventually (hopefully) be moved in a documentation file/page altogether.