diff --git a/src/main.odin b/src/main.odin
index 7a3eef5..1f00267 100644
--- a/src/main.odin
+++ b/src/main.odin
@@ -13,6 +13,10 @@ DATA_FILE :: "res/knapPI_12_500_1000_82.csv"
 NUMBER_OF_ITEMS :: 500
 CAPACITY :: 280785
 POPULATION_SIZE :: 100
+GENERATIONS :: 100
+TOURNAMENT_SIZE :: 3
+CROSSOVER_RATE :: 0.8
+MUTATION_RATE :: 0.01
 
 Item :: struct {
 	profit, weight: int,
@@ -71,7 +75,18 @@ destroy_population :: proc(pop: ^Population) {
 	}
 }
 
-tournament_selection :: proc(pop: ^Population, fitnesses: []int, k: int = 3) -> ^Chromosome {
+evaluate_population :: proc(pop: ^Population) -> (res: [POPULATION_SIZE]int) {
+	for &chrom, i in pop {
+		res[i] = fitness(&chrom)
+	}
+	return
+}
+
+tournament_selection :: proc(
+	pop: ^Population,
+	fitnesses: []int,
+	k: int = TOURNAMENT_SIZE,
+) -> ^Chromosome {
 	best_idx := rand.int_max(POPULATION_SIZE)
 	best_fitness := fitnesses[best_idx]
 
@@ -173,7 +188,7 @@ uniform_crossover :: proc(
 	return
 }
 
-bit_flip_mutation :: proc(chrom: ^Chromosome, mutation_rate: f32 = 0.01) {
+bit_flip_mutation :: proc(chrom: ^Chromosome, mutation_rate: f32 = MUTATION_RATE) {
 	for i in 0 ..< NUMBER_OF_ITEMS {
 		if rand.float32() < mutation_rate {
 			current := bit_array.get(chrom, i)
@@ -240,11 +255,102 @@ elitism_survivor_selection :: proc(
 	return
 }
 
+run_ga :: proc() {
+	population := generate_population()
+	defer destroy_population(&population)
+
+	best_fitness := math.min(int)
+	best_generation := 0
+
+	for gen in 0 ..< GENERATIONS {
+		fitnesses := evaluate_population(&population)
+
+		// log best fitnesses
+		for f, i in fitnesses {
+			if f <= best_fitness {continue}
+			best_fitness = f
+			best_generation = gen
+
+			chrom := &population[i]
+			tot_profit, tot_weight := 0, 0
+			for idx in 0 ..< bit_array.len(chrom) {
+				if !bit_array.get(chrom, idx) {continue}
+				tot_profit += items[idx].profit
+				tot_weight += items[idx].weight
+			}
+			fmt.printfln(
+				"gen %d: best fitness=%d, profit=%d (capacity=%d)",
+				gen,
+				f,
+				tot_profit,
+				tot_weight,
+				CAPACITY,
+			)
+		}
+
+		// create offspring
+		offspring: Population
+		for i := 0; i < POPULATION_SIZE; i += 2 {
+			parent1 := tournament_selection(&population, fitnesses[:])
+			parent2 := tournament_selection(&population, fitnesses[:])
+
+			child1, child2: Chromosome
+			if rand.float32() < CROSSOVER_RATE {
+				child1, child2 = single_point_crossover(parent1, parent2)
+			} else {
+				child1 = bit_array.create(NUMBER_OF_ITEMS)^
+				child2 = bit_array.create(NUMBER_OF_ITEMS)^
+				for j in 0 ..< NUMBER_OF_ITEMS {
+					bit_array.set(&child1, j, bit_array.get(parent1, j))
+					bit_array.set(&child2, j, bit_array.get(parent2, j))
+				}
+			}
+
+			bit_flip_mutation(&child1)
+			bit_flip_mutation(&child2)
+
+			offspring[i] = child1
+			if i + 1 < POPULATION_SIZE {
+				offspring[i + 1] = child2
+			}
+		}
+
+		// survivor selection
+		offspring_fitnesses := evaluate_population(&offspring)
+		new_population := elitism_survivor_selection(
+			&population,
+			&offspring,
+			fitnesses[:],
+			offspring_fitnesses[:],
+		)
+
+		// clean-up
+		destroy_population(&population)
+		destroy_population(&offspring)
+
+		population = new_population
+	}
+
+	fmt.println()
+	fmt.printfln("final best: fitness=%d (found at generation %d)", best_fitness, best_generation)
+}
+
 main :: proc() {
 	items, ok := read_data(DATA_FILE)
 	if !ok {
 		fmt.eprintln("failed to read data from", DATA_FILE)
 		return
 	}
-	fmt.println(items)
+
+	fmt.println("running genetic algorithm for binary knapsack problem.")
+	fmt.printfln(
+		"items: %d, capacity: %d, population: %d, generations: %d",
+		NUMBER_OF_ITEMS,
+		CAPACITY,
+		POPULATION_SIZE,
+		GENERATIONS,
+	)
+	fmt.println()
+
+	run_ga()
 }