vvkin / knapsack-genetic

Genetic algorithm to solve Knapsack problem. Implemented with Python (fully numpy)

Genetic algorithm (Knapsack problem)

Python script to solve Knapsack problem (heuristic). Dynamic programming implementation to check solution precision (exact).

Dependencies

• numpy

Get started

• git clone https://github.com/vvkin/knapsack-genetic/
• cd knapsack-genetic
• pipenv install
• pipenv shell
• pytnon generate_data.py 100 100 int 50 50 data
• python knapsack_genetic.py data 10000

Usage

generate_data.py

Script to generate random data for knapsack problem.
It accepts 6 parameters.
python generate_data.py [items_number] [knapsack_capacity] [items_type] [max_weight] [max_cost] [file_name]

• items_number — integer value. Number of items for problem
• knapsack_capacity — float value. Maximal weight that knapsack can contain
• max_weight — float value. Maximal weight that can be among items, weight ∈ [1, max_weight]
• max_cost — float value. Maximal cost than can be among items, cost ∈ [2, max_cost]
• file_name — string value. File name to write generated data

knapsack_dp.py

Dynamic algorithm to solve Knapsack problem. Can solve problem only with integer weights and costs.
It accepts 1 paratemer.
python knapsack_dp.py [file_name]

• file_name — string value. File name with input data

knapsack_genetic.py

Python script to solve Knapsack problem using genetic algorithm. The algorithm does not guarantee an exact solution, it's heuristic.
It accepts 2 parameters.
python knapsack_genetic.py [file_name] [iterations_number]

• file_name — string value. File name with input data
• iterations_number — integer value. Number of iterations in genetic algorithm

Contributors

Vadym Kinchur, vvkin