shamilton/GeneticAlgo
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

working through compile bugs

Browse Source
This commit is contained in:
Seth Hamilton
2025-08-21 00:41:51 -05:00
parent 3265f045d1
commit 3a901a0a40
3 changed files with 215 additions and 167 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
inc/genetic.h
View File

@@ -2,8 +2,7 @@
namespace genetic {
template <class T> struct ReadonlySpan;
template <class T> struct Span;
template <class T> struct Array;
template <class T> struct Stats;
template <class T> struct Strategy;
@@ -16,7 +15,8 @@ template <class T> struct Strategy {
// before accessing/locking the work queue again.
int num_cells; // Size of the population pool
int num_generations; // Number of times (epochs) to run the algorithm
bool test_all; // Sets whether or not every cell is tested every generation
bool test_all; // Sets whether or not every cell's fitness is evaluated every
// generation
float test_chance; // Chance to test any given cell's fitness. Relevant only
// if test_all is false.
bool enable_crossover; // Cells that score well in the evaluation stage
@@ -25,11 +25,17 @@ template <class T> struct Strategy {
float crossover_mutation_chance; // Chance to mutate a child cell
int crossover_parent_num; // Number of unique high-scoring parents in a
// crossover call.
int crossover_parent_stride; // Number of parents to skip over when moving to
// the next set of parents. A stride of 1 would
// produce maximum overlap because the set of
// parents would only change by one every
// crossover.
int crossover_children_num; // Number of children to expect the user to
// produce in the crossover function.
bool enable_mutation; // Cells may be mutated
// before fitness evaluation
float mutation_chance; // Chance to mutate cells before fitness evaluation
float mutation_chance; // Chance for any given cell to be mutated cells during
// the mutation
uint64_t rand_seed;
bool higher_fitness_is_better; // Sets whether or not to consider higher
// fitness values better or worse. Set this to
@@ -38,16 +44,16 @@ template <class T> struct Strategy {
// User defined functions
T (*make_default_cell)();
void (*mutate)(T &cell_to_modify);
void (*crossover)(const Span<T *> parents, const Span<T *> out_children);
void (*crossover)(const Array<T *> parents, const Array<T *> out_children);
float (*fitness)(const T &cell);
};
template <class T> struct Stats {
std::vector<T> best_cell;
std::vector<float> average_fitness;
std::vector<float> best_cell_fitness;
};
template <class T> struct Span {
template <class T> struct Array {
T *_data;
int len;

178
src/genetic.cpp
View File

@@ -1,9 +1,14 @@
#include "genetic.h"
#include "pthread.h"
#include <algorithm>
#include <cstdint>
#include <cstdlib>
#include <optional>
#include <variant>
#include <vector>
#include "genetic.h"
#include "pthread.h"
#include "rand.h"
#define NUM_QUEUE_RETRIES 10
using namespace std;
@@ -19,23 +24,27 @@ template <class... Ts> overload(Ts...) -> overload<Ts...>;
namespace genetic {
template <class T> struct CellEntry {
template <class T> struct cell_entry {
float score;
T *cell;
bool stale;
};
template <class T> struct CrossoverJob {
Span<CellEntry<T> *> &parents;
Span<CellEntry<T> *> &children_out;
template <class T> struct crossover_job {
Array<cell_entry<T> *> &parents;
Array<cell_entry<T> *> &children_out;
};
template <class T> struct FitnessJob {
CellEntry<T> *cell_entry;
template <class T> struct fitness_job {
cell_entry<T> *cell_entry;
};
template <class T> struct WorkQueue {
variant<CrossoverJob<T>, FitnessJob<T>> *jobs;
template <class T> struct mutate_job {
cell_entry<T> *cell_entry;
};
template <class T> struct work_queue {
variant<crossover_job<T>, fitness_job<T>, mutate_job<T>> *jobs;
int len;
int read_i;
int write_i;
@@ -49,9 +58,9 @@ template <class T> struct WorkQueue {
pthread_cond_t jobs_available_cond;
};
template <class T> WorkQueue<T> make_work_queue(int len) {
return {.jobs = (variant<FitnessJob<T>, CrossoverJob<T>> *)malloc(
sizeof(variant<FitnessJob<T>, CrossoverJob<T>>) * len),
template <class T> work_queue<T> make_work_queue(int len) {
return {.jobs = (variant<fitness_job<T>, crossover_job<T>> *)malloc(
sizeof(variant<fitness_job<T>, crossover_job<T>>) * len),
.len = len,
.read_i = 0,
.write_i = 0,
@@ -63,19 +72,19 @@ template <class T> WorkQueue<T> make_work_queue(int len) {
.jobs_available_cond = PTHREAD_COND_INITIALIZER};
}
template <class T> struct JobBatch {
ReadonlySpan<variant<CrossoverJob<T>, FitnessJob<T>>> jobs;
template <class T> struct job_batch {
Array<variant<crossover_job<T>, fitness_job<T>>> jobs;
bool gen_complete;
};
template <class T>
optional<JobBatch<T>> get_job_batch(WorkQueue<T> &queue, int batch_size,
optional<job_batch<T>> get_job_batch(work_queue<T> &queue, int batch_size,
bool *stop_flag) {
while (true) {
for (int i = 0; i < NUM_QUEUE_RETRIES; i++) {
if (queue.read_i < queue.write_i &&
pthread_mutex_trylock(&queue.data_mutex)) {
JobBatch<T> res;
job_batch<T> res;
res.jobs._data = &queue._jobs[queue.read_i];
int span_size = min(batch_size, queue.write_i - queue.read_i);
res.jobs.len = span_size;
@@ -94,27 +103,41 @@ optional<JobBatch<T>> get_job_batch(WorkQueue<T> &queue, int batch_size,
}
}
template <class T> struct WorkerThreadArgs {
template <class T> struct worker_thread_args {
Strategy<T> &strat;
WorkQueue<T> &queue;
work_queue<T> &queue;
bool *stop_flag;
};
template <class T> void do_crossover_job(CrossoverJob<T> cj) {}
template <class T> void *worker(void *args) {
WorkerThreadArgs<T> *work_args = (WorkerThreadArgs<T> *)args;
worker_thread_args<T> *work_args = (worker_thread_args<T> *)args;
Strategy<T> &strat = work_args->strat;
WorkQueue<T> &queue = work_args->queue;
work_queue<T> &queue = work_args->queue;
bool *stop_flag = work_args->stop_flag;
auto JobDispatcher = overload{
[strat](FitnessJob<T> fj) {
fj.cell_entry->result_out = strat.fitness(*(fj.cell_entry->cell));
fj.cell_entry->stale = true;
auto job_dispatcher = overload{
[strat](mutate_job<T> mj) {
strat.mutate(*mj.cell_entry->cell);
mj.cell_entry->stale = true;
},
[strat](CrossoverJob<T> cj) {
strat.crossover(cj.parents, cj.children_out);
[strat](fitness_job<T> fj) {
fj.cell_entry->score = strat.fitness(*fj.cell_entry->cell);
fj.cell_entry->stale = false;
},
[strat](crossover_job<T> cj) {
Array<T *> parent_cells, child_cells;
parent_cells = {(T **)malloc(sizeof(T *) * cj.parents.len),
cj.parents.len};
child_cells = {(T **)malloc(sizeof(T *) * cj.children_out.len),
cj.children_out.len};
for (int i = 0; i < cj.parents.len; i++) {
parent_cells[i] = cj.parents[i].cell;
}
for (int i = 0; i < cj.children_out.len; i++) {
child_cells[i] = cj.children_out[i].cell;
cj.children_out[i].stale = true;
}
strat.crossover(parent_cells, child_cells);
},
};
@@ -125,7 +148,7 @@ template <class T> void *worker(void *args) {
// Do the actual work
for (int i = 0; i < batch->jobs.len; i++) {
visit(JobDispatcher, batch->jobs[i]);
visit(job_dispatcher, batch->jobs[i]);
}
if (batch->gen_complete) {
@@ -136,20 +159,24 @@ template <class T> void *worker(void *args) {
template <class T> Stats<T> run(Strategy<T> strat) {
Stats<T> stats;
WorkQueue<T> work_queue = make_work_queue<T>(strat.num_cells);
// The work queue is what all the worker threads will checking
// for jobs
work_queue<T> queue = make_work_queue<T>(strat.num_cells);
// The actual cells. Woo!
T cells[strat.num_cells];
// Using a vector so I can use the make_heap, push_heap, etc.
vector<CellEntry<T>> cell_queue;
vector<cell_entry<T>> cell_queue;
for (int i = 0; i < strat.num_cells; i++) {
cells[i] = strat.make_default_cell();
cell_queue.push_back({0, &cells[i], true});
}
bool stop_flag = false;
WorkerThreadArgs<T> args = {
.strat = strat, .queue = work_queue, .stop_flag = &stop_flag};
worker_thread_args<T> args = {
.strat = strat, .queue = queue, .stop_flag = &stop_flag};
// spawn worker threads
pthread_t threads[strat.num_threads];
@@ -157,28 +184,95 @@ template <class T> Stats<T> run(Strategy<T> strat) {
pthread_create(&threads[i], NULL, worker<T>, (void *)args);
}
for (int i = 0; i < strat.num_generations; i++) {
// generate fitness jobs
if (strat.test_all) {
uint64_t rand_state = strat.rand_seed;
} else {
for (int i = 0; i < strat.num_generations; i++) {
// Mutate some random cells in the population
for (int i = 0; i < cell_queue.size(); i++) {
if (abs(norm_rand(rand_state)) < strat.mutation_chance) {
queue.jobs[queue.write_i] = mutate_job<T>{&cell_queue[i]};
queue.write_i++;
}
}
pthread_cond_broadcast(&queue.jobs_available_cond);
// Potential issue here where mutations aren't done computing and fitness
// jobs begin. maybe need to gate this.
// Generate fitness jobs
for (int i = 0; i < cell_queue.size(); i++) {
if (cell_queue[i].stale &&
(strat.test_all || abs(norm_rand(rand_state)) < strat.test_chance)) {
queue.jobs[queue.write_i] = fitness_job<T>{&cell_queue[i]};
queue.write_i++;
}
pthread_cond_broadcast(&queue.jobs_available_cond);
}
queue.done_writing = true;
// wait for fitness jobs to complete
// sort cells on performance
pthread_mutex_lock(&queue.gen_complete_mutex);
// Before going to sleep, do a quick check to see if the fitness jobs are
// already complete.
pthread_mutex_lock(&queue.data_mutex);
bool already_complete = queue.read_i != queue.write_i;
pthread_mutex_unlock(&queue.data_mutex);
if (already_complete) {
pthread_mutex_unlock(&queue.gen_complete_mutex);
} else {
pthread_cond_wait(&queue.gen_complete_cond, &queue.gen_complete_mutex);
}
// Sort cells on performance
std::sort(cell_queue.begin(), cell_queue.end(),
[strat](cell_entry<T> a, cell_entry<T> b) {
return strat.higher_fitness_is_better ? a > b : a < b;
});
printf("Top Score: %f\n", cell_queue[0].score);
if (!strat.enable_crossover)
continue;
// generate crossover jobs
// dear god. forgive me father
queue.write_i = 0;
queue.read_i = 0;
int count = 0;
int n_par = strat.crossover_parent_num;
int n_child = strat.crossover_children_num;
int child_i = cell_queue.size() - 1;
int par_i = 0;
while (child_i - par_i <= n_par + n_child) {
Array<cell_entry<T> *> parents = {
(cell_entry<T> **)malloc(sizeof(cell_entry<T> *) * n_par), n_par};
Array<cell_entry<T> *> children = {
(cell_entry<T> **)malloc(sizeof(cell_entry<T> *) * n_child), n_child};
for (; par_i < par_i + n_par; par_i++) {
parents[i] = cell_queue[par_i];
}
for (; child_i > child_i - n_child; child_i--) {
children[i] = cell_queue[child_i];
}
queue.jobs[queue.write_i] = crossover_job<T>{parents, children};
par_i += strat.crossover_parent_stride;
child_i += strat.crossover_children_stride;
}
}
// stop worker threads
stop_flag = true;
pthread_cond_broadcast(work_queue.jobs_available_cond);
pthread_cond_broadcast(&queue.jobs_available_cond);
for (int i = 0; i < strat.num_threads; i++) {
pthread_join(threads[i], NULL);
}
}
template <class T> T &Span<T>::operator[](int i) {
assert(i >= 0 && i < len);
template <class T> T &Array<T>::operator[](int i) {
return _data[i];
}

180
src/main.cpp
View File

@@ -1,133 +1,81 @@
#include <algorithm>
#include <cassert>
#include <cstdint>
#include <cstdlib>
#include <iostream>
#include <vector>
#include "genetic.h"
#include "rand.h"
#define MUTATION_CHANCE 1.0
using namespace genetic;
float norm_rand() { return (float)rand() / RAND_MAX; }
const int len = 10;
const float max_float = 9999.9f;
static uint64_t seed = 12;
static float num_mutate_chance = 0.5;
static int num_parents = 2;
static int num_children = 2;
enum class ConstraintType {
PRODUCT = 0,
SUM = 1,
INDEX_EQ = 2,
};
struct Constraint {
ConstraintType type;
int optional_i;
float value;
};
static std::vector<Constraint> constraints;
static int target_sum = 200;
static int target_product = 300;
struct Cell {
int n;
float *params;
};
Cell make_cell(int num_params) {
Cell res = {num_params, (float *)malloc(num_params * sizeof(float))};
for (int i = 0; i < num_params; i++) {
res.params[i] = 0;
Array<float> make_new_arr() {
Array<float> arr = { (float*)malloc(sizeof(float)*len), len };
for (int i = 0; i < arr.len; i++) {
arr[i] = norm_rand(seed) * max_float;
}
return res;
return arr;
}
float get_cell_err(const Cell &a) {
float total_diff = 0;
for (auto c : constraints) {
switch (c.type) {
case ConstraintType::SUM: {
void mutate(Array<float> &arr_to_mutate) {
for (int i = 0; i < len; i++) {
if (norm_rand(seed) < num_mutate_chance) {
arr_to_mutate[i] = norm_rand(seed) * max_float;
}
}
}
void crossover(const Array<Array<float>*> parents, const Array<Array<float> *> out_children) {
for (int i = 0; i < len; i++) {
(*out_children._data[0])[i] = i < len/2 ? (*parents._data[0])[i] : (*parents._data[1])[i];
(*out_children._data[1])[i] = i < len/2 ? (*parents._data[1])[i] : (*parents._data[0])[i];
}
}
// norm_rand can go negative. fix in genetic.cpp
// child stride doesn't make sense. Should always skip over child num
float fitness(const Array<float> &cell) {
float sum = 0;
for (int i = 0; i < a.n; i++) {
sum += a.params[i];
float product = 1;
for (int i = 0; i < cell.len; i++) {
sum += cell._data[i];
product *= cell._data[i];
}
total_diff += abs(c.value - sum);
break;
}
case ConstraintType::PRODUCT: {
float prod = 1;
for (int i = 0; i < a.n; i++) {
prod *= a.params[i];
}
total_diff += abs(c.value - prod);
break;
}
case ConstraintType::INDEX_EQ: {
assert(c.optional_i < a.n);
total_diff += abs(c.value - a.params[c.optional_i]);
break;
}
}
}
return total_diff;
}
bool operator<(const Cell &a, const Cell &b) {
assert(a.n == b.n);
return get_cell_err(a) < get_cell_err(b);
}
void combine_cells(const Cell &a, const Cell &b, Cell *child) {
bool a_first = norm_rand() > 0.5f;
for (int i = 0; i < a.n; i++) {
float offset = norm_rand() * 10;
float roll = norm_rand();
if (a_first) {
child->params[i] = (i < a.n / 2 ? a.params[i] : b.params[i]) +
(roll > 0.5 ? offset : -offset);
} else {
child->params[i] = (i < a.n / 2 ? b.params[i] : a.params[i]) +
(roll > 0.5 ? offset : -offset);
}
}
float r = norm_rand();
child->params[(int)r * (a.n - 1)] = r * 100.0;
return abs(sum - target_sum) + abs(product - target_product);
}
int main(int argc, char **argv) {
int num_params, num_cells, num_generations, num_constraints = 0;
std::cin >> num_params >> num_cells >> num_generations >> num_constraints;
Strategy<Array<float>> strat {
.num_threads = 1,
.batch_size = 1,
.num_cells = 10,
.num_generations = 10,
.test_all = true,
.test_chance = 0.0, // doesn't matter
.enable_crossover = true,
.enable_crossover_mutation = true,
.crossover_mutation_chance = 0.6f,
.crossover_parent_num = 2,
.crossover_parent_stride = 1,
.crossover_children_num = 2,
.enable_mutation = true,
.mutation_chance = 0.8,
.rand_seed = seed,
.higher_fitness_is_better = false,
.make_default_cell=make_new_arr,
.mutate=mutate,
.crossover=crossover,
.fitness=fitness
};
std::cout << num_params << " " << num_cells << " " << num_generations << " "
<< num_constraints << std::endl;
for (int i = 0; i < num_constraints; i++) {
int type_in, optional_i = 0;
float value;
std::cin >> type_in >> value;
ConstraintType type = static_cast<ConstraintType>(type_in);
if (type == ConstraintType::INDEX_EQ) {
std::cin >> optional_i;
}
constraints.push_back({type, optional_i, value});
}
std::vector<Cell> cells;
for (int i = 0; i < num_cells; i++) {
cells.push_back(make_cell(num_params));
}
for (int i = 0; i < num_generations; i++) {
std::sort(cells.begin(), cells.end());
for (int j = 0; j < num_cells / 2; j++) {
combine_cells(cells[j], cells[j + 1], &cells[num_cells / 2 + j]);
}
if (i % 1000 == 0) {
std::cout << i << "\t" << get_cell_err(cells[0]) << std::endl;
}
}
std::cout << "Final Answer: ";
float sum = 0;
float product = 1;
for (int i = 0; i < cells[0].n; i++) {
std::cout << cells[0].params[i] << " ";
sum += cells[0].params[i];
product *= cells[0].params[i];
}
std::cout << std::endl;
std::cout << "Sum: " << sum << std::endl;
std::cout << "Product: " << product << std::endl;
auto res = run(strat);
}