Flemme Olympique - Interactive Operations Research Workshops for Terra Numerica

The goal of flemolymp is to provide a hands-on introduction to the subset sum problem in operations research.

Participants tackle computational challenges while creating Olympic-style highlight videos, combining algorithmic thinking with creative storytelling. This activity is designed for Terra Numerica, offering an engaging way to explore optimization concepts in a playful and interactive setting.

Installation

You can install the development version of flemolymp from GitHub with:

# install.packages("pak")
pak::pak("arnaud-m/flemolymp")

Example

This is a basic example which shows you how to solve a subset sum problem with greedy search and dynamic programming:

library(flemolymp)
# Example input: list of integer vectors representing partition instances
inputs <- list(
  c(11, 8, 7, 5, 2, 1),
  c(14, 13, 11, 7, 5, 3)
)

# Compute summary statistics for the instances
x <- SummaryInstances(inputs)

# Inspect the summary statistics
print(x$outputs)

# Solve the first instance with the greedy algorithm
solutionGreedy <- GreedySearchSSP(
  sizes = inputs[[1]],
  capacity = HalfSum(inputs[[1]])
)

cat("Greedy solution for instance 1:\n")
print(solutionGreedy)
print(inputs[[1]][sapply(solutionGreedy, isTRUE)])

# Solve the second instance with dynamic programming
reachedDP <- DynamicProgrammingSSP(
  sizes = inputs[[2]],
  capacity = HalfSum(inputs[[2]])
)

cat("Dynamic programming solution for instance 2:\n")
solutionDP <- GetSolutionDP(reachedDP, inputs[[2]])
print(solutionDP)
print(inputs[[2]][sapply(solutionDP, isTRUE)])

Licenses

• The R code is licensed under the MIT License. See LICENSE.md for details.
• Pedagogical content (images, text, diagrams) is licensed under CC BY 4.0. See LICENSE_CONTENT.md for details.

Copyright 2026 Université Côte d'Azur. All rights reserved.