🧬 Simple Genetic Algorithm in Python

This project demonstrates how to implement a Genetic Algorithm (GA) from scratch in Python — a fun way to mimic natural selection and evolve solutions. The goal is to guess a target string using random populations, fitness evaluation, selection, crossover, mutation, and population regeneration.

Why Python? Because I love it.
Hate Python? That's your problem 😛

---

📦 Requirements

Make sure to install the only dependency:

pip install numpy

---

📊 Genetic Algorithm Flow

A Genetic Algorithm simulates the process of natural evolution to solve optimization problems. The main components are:

1. Initialization – Create a random population of possible solutions (strings).
2. Selection – Pick the fittest individuals from the population.
3. Crossover – Combine parts of parents to produce offspring.
4. Mutation – Randomly tweak offspring to introduce diversity.
5. Evaluation – Stop if the target is reached or continue to next generation.
6. Regeneration – Replace the worst individuals with new ones.

🧭 Flowchart Overview

---

🔧 Step-by-Step Implementation

1. Initialize Population

A population is a collection of individuals, called genes. Each gene is a random string.

def create_gen(panjang_target):
    random_number = np.random.randint(32, 126, size=panjang_target)
    return ''.join([chr(i) for i in random_number])

Each gene’s fitness is evaluated against the target string.

def calculate_fitness(gen, target, panjang_target):
    matches = sum([1 for i in range(panjang_target) if gen[i] == target[i]])
    return (matches / panjang_target) * 100

All genes are stored with their fitness in a dictionary:

def create_population(target, max_population, panjang_target):
    populasi = {}
    for _ in range(max_population):
        gen = create_gen(panjang_target)
        populasi[gen] = calculate_fitness(gen, target, panjang_target)
    return populasi

---

2. Selection

Choose the two best genes (with the highest fitness values):

def selection(populasi):
    pop = dict(populasi)
    parent = {}
    for _ in range(2):
        gen = max(pop, key=pop.get)
        parent[gen] = pop[gen]
        del pop[gen]
    return parent

---

3. Crossover

Mix the selected parents to create two children by swapping halves:

def crossover(parent, target, panjang_target):
    cp = round(panjang_target / 2)
    p = list(parent)
    return {
        p[0][:cp] + p[1][cp:]: calculate_fitness(p[0][:cp] + p[1][cp:], target, panjang_target),
        p[1][:cp] + p[0][cp:]: calculate_fitness(p[1][:cp] + p[0][cp:], target, panjang_target)
    }

---

4. Mutation

Introduce diversity by randomly changing characters with a certain mutation rate:

def mutation(child, target, mutation_rate, panjang_target):
    mutant = {}
    for c in child:
        data = list(c)
        for j in range(len(data)):
            if np.random.rand() <= mutation_rate:
                data[j] = chr(np.random.randint(32, 126))
        gen = ''.join(data)
        mutant[gen] = calculate_fitness(gen, target, panjang_target)
    return mutant

---

5. Evaluation and Stopping Criteria

Check if the best individual has reached the target (100% fitness):

if bestfitness(mutant) >= 100:
    break

---

6. Regeneration

Replace the weakest genes in the population with the newly mutated ones:

def regeneration(mutant, populasi):
    for _ in mutant:
        bad_gen = min(populasi, key=populasi.get)
        del populasi[bad_gen]
    populasi.update(mutant)
    return populasi

---

🧪 Example Output

Target Word : Hello World!
Max Population : 10
Mutation Rate : 0.2
----------------------------------------------
The Best        Fitness     Time
----------------------------------------------
HgUewI]W*f`     15.38%      0:00:00.003001
Helpo Worsd!    84.61%      0:00:01.003120
Hello World!    100.0%      0:00:01.994801

---

▶️ Run the Program

python genetic_algorithm.py

Edit the config at the top of the file:

target = 'Hello World!'
max_population = 10
mutation_rate = 0.2

---

📜 License

This project is licensed under the MIT License.

---

🙋‍♂️ Author

Ardy Seto Priambodo 📧 2black0@gmail.com

---

📚 References

• Goldberg, D. E., Genetic Algorithms in Search, Optimization, and Machine Learning
• Mitchell, M., An Introduction to Genetic Algorithms