Rabbit Population Growth Simulation

Stochastic simulation of rabbit population dynamics in C, comparing an idealized Fibonacci growth model with a realistic model incorporating biological parameters (mortality, reproduction, predation, disease).

University project — L2 Computer Science, Université Clermont-Auvergne (2025)

Overview

This project implements two models:

1. Fibonacci Model simple exponential growth with no mortality
2. Stochastic Model realistic simulation with:
   • Age-dependent survival rates (35% juvenile, 60% adult, declining after 10 years)
   • Predator and disease mortality (5% base + 1% per million individuals)
   • Gaussian-distributed litter sizes (3-9 kits) and litters per year (3-9)
   • Sexual maturity between 5-8 months (normal distribution)
   • Individual tracking via doubly-linked list for O(1) insertion/deletion
   • Mersenne Twister (MT19937-64) random number generator

Key results (20 simulations, 140 months each):

Metric	Value
Mean population	3,477,703
Standard deviation	2,319,671
95% confidence interval	[2,392,076 4,563,330]
Range	0 6,784,794

Project Structure

.
 fibonacci-model/     Fibonacci model (simple)
    question1.c
    mt19937-64.c
    mt64.h
 stochastic-model/    Stochastic model (realistic)
    main.c            Entry point
    rabbit.c/.h       Rabbit data structures and management
    simu.c/.h         Simulation logic and statistics
    mt19937-64.c      Mersenne Twister RNG
    mt64.h
    simuResult/       CSV output files
 Rapport TP04/        Report (PDF) with graphs and analysis
 Makefile

Build & Run

Requirements: GCC, Make

# Build both models
make all

# Or build individually
make fibonacci
make stochastic

# Run Fibonacci model
./fibonacci-model/fibonacci

# Run stochastic simulation (outputs CSV files to stochastic-model/simuResult/)
./stochastic-model/simulation

# Clean
make clean

Technical Highlights

• Zero memory leaks validated with Valgrind (46.7M allocations, all freed)
• Performance profiled with gprof to identify bottlenecks
• Monte Carlo approach 20 independent runs for statistical validity
• CSV output for data analysis and visualization

Author

• Titouan Mota