Merge branch 'alternate_multithreaded'
This commit is contained in:
379
inc/genetic.h
379
inc/genetic.h
@@ -1,9 +1,10 @@
|
||||
#pragma once
|
||||
|
||||
#include <algorithm>
|
||||
#include <cfloat>
|
||||
#include <cstdlib>
|
||||
#include <vector>
|
||||
|
||||
#include "util.h"
|
||||
#include "sync.h"
|
||||
#include "rand.h"
|
||||
|
||||
@@ -12,21 +13,31 @@ using namespace std;
|
||||
|
||||
namespace genetic {
|
||||
|
||||
template <class T> struct Array;
|
||||
template <class T> struct Stats;
|
||||
template <class T> struct Strategy;
|
||||
struct CellTracker;
|
||||
|
||||
template <class T> Stats<T> run(Strategy<T>);
|
||||
template <class T> T run(Strategy<T>);
|
||||
|
||||
template <class T> struct Strategy {
|
||||
// Number of worker threads that will be evaluating cell fitness
|
||||
int num_threads;
|
||||
|
||||
int batch_size; // Number of cells a worker thread tries to work on in a row
|
||||
// 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
|
||||
// Period of print statements (in seconds)
|
||||
float stats_print_period_s;
|
||||
|
||||
// Size of the population pool per sim thread
|
||||
int num_cells_per_thread;
|
||||
|
||||
// Number of times (epochs) to run the algorithm
|
||||
int num_generations;
|
||||
|
||||
// Each thread will integrate the best globally performing cell
|
||||
bool share_breakthroughs;
|
||||
|
||||
// How many generations to explore before resyncing with the global best
|
||||
int share_breakthrough_gen_period;
|
||||
|
||||
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
|
||||
@@ -59,10 +70,16 @@ template <class T> struct Strategy {
|
||||
};
|
||||
|
||||
template<class T> struct Stats {
|
||||
std::vector<T> best_cell;
|
||||
std::vector<float> best_cell_fitness;
|
||||
TimeSpan setup_time;
|
||||
TimeSpan run_time;
|
||||
DynArray<T> best_cells;
|
||||
DynArray<float> best_cell_fitness;
|
||||
int gen;
|
||||
bool done;
|
||||
DynArray<TimeSpan> gen_time;
|
||||
DynArray<TimeSpan> crossover_time;
|
||||
DynArray<TimeSpan> mutate_time;
|
||||
DynArray<TimeSpan> fitness_time;
|
||||
DynArray<TimeSpan> sorting_time;
|
||||
Mutex m;
|
||||
};
|
||||
|
||||
struct CellTracker {
|
||||
@@ -70,150 +87,81 @@ struct CellTracker {
|
||||
int cellid;
|
||||
};
|
||||
|
||||
template <class T> struct Array {
|
||||
T *data;
|
||||
int len;
|
||||
|
||||
T &operator[](int i) { return data[i]; }
|
||||
};
|
||||
|
||||
template <class T> Array<T> make_array(int len) {
|
||||
return {
|
||||
.data = (T*)malloc(sizeof(T)*len),
|
||||
.len = len
|
||||
};
|
||||
}
|
||||
|
||||
template<class T>
|
||||
struct MutateJob {
|
||||
T* cell;
|
||||
};
|
||||
|
||||
template<class T>
|
||||
struct CrossoverJob {
|
||||
Array<T*> parents;
|
||||
Array<T*> children;
|
||||
};
|
||||
|
||||
template<class T>
|
||||
struct FitnessJob {
|
||||
T* cell;
|
||||
CellTracker* track;
|
||||
};
|
||||
|
||||
enum class JobType {
|
||||
MUTATE,
|
||||
CROSSOVER,
|
||||
FITNESS
|
||||
};
|
||||
|
||||
template<class T>
|
||||
union Job {
|
||||
MutateJob<T> m;
|
||||
CrossoverJob<T> c;
|
||||
FitnessJob<T> f;
|
||||
};
|
||||
|
||||
// Yes. I am aware of variant
|
||||
// For some reason I like this better
|
||||
template<class T>
|
||||
struct TaggedJob {
|
||||
Job<T> data;
|
||||
JobType type;
|
||||
};
|
||||
|
||||
template<class T>
|
||||
struct WorkQueue {
|
||||
Array<TaggedJob<T>> jobs;
|
||||
int read_i, write_i, batch_size;
|
||||
bool done_writing, work_complete, stop; // These catch some edge conditions
|
||||
Mutex m;
|
||||
ConditionVar done;
|
||||
ConditionVar jobs_ready;
|
||||
};
|
||||
|
||||
template<class T>
|
||||
struct WorkerThreadArgs {
|
||||
WorkQueue<T> &q;
|
||||
Strategy<T> &s;
|
||||
Strategy<T> strat;
|
||||
Array<T> cells;
|
||||
Array<CellTracker> trackers;
|
||||
Stats<T> *stats;
|
||||
|
||||
Mutex m;
|
||||
float *best_global_score;
|
||||
T* best_global_cell;
|
||||
};
|
||||
|
||||
template<class T> WorkQueue<T> make_work_queue(int len, int batch_size);
|
||||
template<class T> bool tryget_job_batch(WorkQueue<T> &q, int len, Array<TaggedJob<T>>* out_batch, bool* out_batch_is_end);
|
||||
template<class T> T* _cellp(Array<T> cells, CellTracker tracker) { return &cells[tracker.cellid]; }
|
||||
|
||||
template<class T>
|
||||
DWORD worker(LPVOID args);
|
||||
template <class T> DWORD worker(LPVOID args) {
|
||||
// Unpack everything...
|
||||
WorkerThreadArgs<T>* worker_args = static_cast<WorkerThreadArgs<T>*>(args);
|
||||
Strategy<T> strat = worker_args->strat;
|
||||
Array<T> cells = worker_args->cells;
|
||||
Array<CellTracker> trackers = worker_args->trackers;
|
||||
Stats<T> &stats = *worker_args->stats;
|
||||
float* best_global_score = worker_args->best_global_score;
|
||||
T* best_global_cell = worker_args->best_global_cell;
|
||||
Mutex best_m = worker_args->m;
|
||||
|
||||
template <class T> Stats<T> run(Strategy<T> strat) {
|
||||
Stats<T> stats;
|
||||
// Prepare crossover operations as these will be the same every time except
|
||||
// for the exact cell pointers
|
||||
int npar = strat.crossover_parent_num;
|
||||
int nchild = strat.crossover_children_num;
|
||||
Array<T*> parents = make_array<T*>(npar);
|
||||
Array<T*> children = make_array<T*>(nchild);
|
||||
|
||||
// ************* SETUP **************
|
||||
TimeSpan start_setup = now();
|
||||
bool gt = strat.higher_fitness_is_better; // Writing strat.higher... is annoying
|
||||
|
||||
// Create cells
|
||||
Array<T> cells = make_array<T>(strat.num_cells);
|
||||
for (int i = 0; i < cells.len; i++) cells[i] = strat.make_default_cell();
|
||||
// printf("Core: %d\n", get_affinity());
|
||||
|
||||
// Create cell trackers
|
||||
Array<CellTracker> trackers = make_array<CellTracker>(strat.num_cells);
|
||||
for (int i = 0; i < trackers.len; i++) trackers[i] = { .score=0, .cellid=i };
|
||||
TimeSpan start, diff, gen_start;
|
||||
while(stats.gen < strat.num_generations) {
|
||||
gen_start = now();
|
||||
|
||||
// Create work queue
|
||||
// Worst case size is every cell mutated, crossed, and evaluated...? Not quite, but 3x should be upper bound
|
||||
WorkQueue<T> q = make_work_queue<T>(3*strat.num_cells, strat.batch_size);
|
||||
WorkerThreadArgs<T> args = {q, strat};
|
||||
|
||||
// Create worker threads
|
||||
Thread *threads = (Thread*)malloc(sizeof(Thread*)*strat.num_threads);
|
||||
for (int i = 0; i < strat.num_threads; i++) {
|
||||
threads[i] = make_thread(worker<T>, &args);
|
||||
}
|
||||
|
||||
stats.setup_time = now() - start_setup;
|
||||
|
||||
// *********** ALGORITHM ************
|
||||
TimeSpan start_algo = now();
|
||||
for (int gen = 0; gen < strat.num_generations; gen++) {
|
||||
// Reset work queue
|
||||
lock(q.m);
|
||||
q.read_i = 0;
|
||||
q.write_i = 0;
|
||||
q.work_complete = false;
|
||||
q.done_writing = false;
|
||||
unlock(q.m);
|
||||
|
||||
// 1. mutate
|
||||
for (int i = 0; i < trackers.len; i++) {
|
||||
if (abs(norm_rand(strat.rand_seed)) < strat.mutation_chance) {
|
||||
MutateJob<T> mj = {&cells[trackers[i].cellid]};
|
||||
TaggedJob<T> job;
|
||||
job.data.m = mj;
|
||||
job.type=JobType::MUTATE;
|
||||
q.jobs[q.write_i++] = job;
|
||||
// 0. Share/Integrate global breakthrough
|
||||
if (strat.share_breakthroughs && (stats.gen + get_affinity()) % strat.share_breakthrough_gen_period) {
|
||||
lock(best_m);
|
||||
if (better(gt, front(trackers).score, *best_global_score) != *best_global_score) {
|
||||
// Share
|
||||
*best_global_cell = *_cellp(cells, trackers[0]);
|
||||
*best_global_score = trackers[0].score;
|
||||
|
||||
} else {
|
||||
// Integrate
|
||||
*_cellp(cells, trackers[0]) = *best_global_cell;
|
||||
trackers[0].score = *best_global_score;
|
||||
}
|
||||
unlock(best_m);
|
||||
}
|
||||
wake_all(q.jobs_ready); // There are available jobs for the worker threads!
|
||||
|
||||
// 2. crossover
|
||||
// 1. crossover
|
||||
start = now();
|
||||
if (strat.enable_crossover) {
|
||||
int npar = strat.crossover_parent_num;
|
||||
int nchild = strat.crossover_children_num;
|
||||
|
||||
int parent_end = npar;
|
||||
int child_begin = trackers.len-nchild;
|
||||
while (parent_end <= child_begin) {
|
||||
|
||||
// TODO: Variable size arrays please. This is rediculous.
|
||||
Array<T*> parents = make_array<T*>(npar);
|
||||
Array<T*> children = make_array<T*>(nchild);
|
||||
// Get pointers to all the parent cells
|
||||
for (int i = parent_end-npar; i < parent_end; i++) {
|
||||
parents[i - (parent_end-npar)] = &cells[trackers[i].cellid];
|
||||
T* cell = _cellp(cells, trackers[i]);
|
||||
assert(cell != NULL);
|
||||
parents[i - (parent_end-npar)] = cell;
|
||||
}
|
||||
|
||||
// Get pointers to all the child cells (these will be overwritten)
|
||||
for (int i = child_begin; i < child_begin+nchild; i++) {
|
||||
children[i-child_begin] = &cells[trackers[i].cellid];
|
||||
T* cell = _cellp(cells, trackers[i]);
|
||||
assert(cell != NULL);
|
||||
children[i-child_begin] = cell;
|
||||
}
|
||||
CrossoverJob<T> cj = {parents, children};
|
||||
TaggedJob<T> job;
|
||||
@@ -223,10 +171,25 @@ template <class T> Stats<T> run(Strategy<T> strat) {
|
||||
parent_end += strat.crossover_parent_stride;
|
||||
child_begin -= nchild;
|
||||
}
|
||||
wake_all(q.jobs_ready); // There are available jobs for the worker threads!
|
||||
}
|
||||
lock(stats.m);
|
||||
append(stats.crossover_time, now() - start);
|
||||
unlock(stats.m);
|
||||
|
||||
|
||||
// 2. mutate
|
||||
start = now();
|
||||
for (int i = 0; i < trackers.len; i++) {
|
||||
if (abs(norm_rand(strat.rand_seed)) < strat.mutation_chance) {
|
||||
strat.mutate(cells[trackers[i].cellid]);
|
||||
}
|
||||
}
|
||||
lock(stats.m);
|
||||
append(stats.mutate_time, now() - start);
|
||||
unlock(stats.m);
|
||||
|
||||
// 3. evaluate
|
||||
start = now();
|
||||
if (strat.test_all) {
|
||||
for (int i = 0; i < trackers.len; i++) {
|
||||
FitnessJob<T> fj = {&cells[trackers[i].cellid], &trackers[i]};
|
||||
@@ -251,29 +214,147 @@ template <class T> Stats<T> run(Strategy<T> strat) {
|
||||
q.done_writing = true;
|
||||
unlock(q.m);
|
||||
}
|
||||
wake_all(q.jobs_ready);
|
||||
|
||||
// Wait until the work is finished
|
||||
lock(q.m);
|
||||
if (!q.work_complete)
|
||||
wait(q.done, q.m, infinite_ts);
|
||||
unlock(q.m);
|
||||
lock(stats.m);
|
||||
append(stats.fitness_time, now() - start);
|
||||
unlock(stats.m);
|
||||
|
||||
// 4. sort
|
||||
std::sort(&trackers[0], &trackers[trackers.len-1], [strat](CellTracker &a, CellTracker &b){ return strat.higher_fitness_is_better ? a.score > b.score : a.score < b.score; });
|
||||
start = now();
|
||||
std::sort(&trackers[0], &trackers[trackers.len-1], [strat](CellTracker &a, CellTracker &b){ return better(strat.higher_fitness_is_better, a.score, b.score) == a.score; });
|
||||
lock(stats.m);
|
||||
append(stats.sorting_time, now() - start);
|
||||
|
||||
printf("Gen: %d, Best Score: %f\n", gen, trackers[0].score);
|
||||
stats.best_cell.push_back(cells[trackers[0].cellid]);
|
||||
stats.best_cell_fitness.push_back(trackers[0].score);
|
||||
append(stats.best_cells, cells[trackers[0].cellid]);
|
||||
append(stats.best_cell_fitness, trackers[0].score);
|
||||
append(stats.gen_time, now() - gen_start);
|
||||
stats.gen++;
|
||||
unlock(stats.m);
|
||||
}
|
||||
stats.done = true;
|
||||
return 0;
|
||||
}
|
||||
|
||||
template <class T> T run(Strategy<T> strat) {
|
||||
Array<Stats<T>> stats = make_array<Stats<T>>(strat.num_threads);
|
||||
Array<Thread> threads = make_array<Thread>(strat.num_threads);
|
||||
Array<WorkerThreadArgs<T>> args = make_array<WorkerThreadArgs<T>>(strat.num_threads);
|
||||
|
||||
float best_global_score = strat.higher_fitness_is_better ? FLT_MIN : FLT_MAX;
|
||||
T best_global_cell;
|
||||
|
||||
allow_all_processors();
|
||||
set_affinity(0);
|
||||
|
||||
for (int i = 0; i < strat.num_threads; i++) {
|
||||
stats[i] = {
|
||||
.best_cells=make_dynarray<T>(strat.num_generations),
|
||||
.best_cell_fitness=make_dynarray<float>(strat.num_generations),
|
||||
.gen_time=make_dynarray<TimeSpan>(strat.num_generations),
|
||||
.crossover_time=make_dynarray<TimeSpan>(strat.num_generations),
|
||||
.mutate_time=make_dynarray<TimeSpan>(strat.num_generations),
|
||||
.fitness_time=make_dynarray<TimeSpan>(strat.num_generations),
|
||||
.sorting_time=make_dynarray<TimeSpan>(strat.num_generations),
|
||||
.m=make_mutex()
|
||||
};
|
||||
Array<T> cells = make_array<T>(strat.num_threads*strat.num_cells_per_thread);
|
||||
Array<CellTracker> trackers = make_array<CellTracker>(strat.num_cells_per_thread);
|
||||
for (int i = 0; i < strat.num_cells_per_thread; i++) {
|
||||
cells[i] = strat.make_default_cell();
|
||||
trackers[i] = {0, i};
|
||||
}
|
||||
|
||||
args[i].strat=strat;
|
||||
args[i].cells=cells;
|
||||
args[i].trackers=trackers;
|
||||
args[i].stats=&stats[i];
|
||||
args[i].best_global_score=&best_global_score;
|
||||
args[i].best_global_cell=&best_global_cell;
|
||||
args[i].m = make_mutex();
|
||||
|
||||
threads[i] = make_thread(worker<T>, &args[i], i+1);
|
||||
}
|
||||
|
||||
q.stop = true;
|
||||
wake_all(q.jobs_ready);
|
||||
// TODO: join all threads
|
||||
// We are the stats thread
|
||||
bool complete = false;
|
||||
while (!complete) {
|
||||
sleep(from_s(strat.stats_print_period_s));
|
||||
|
||||
// TODO: There's some data freeing that should really be done here
|
||||
stats.run_time = now() - start_algo;
|
||||
return stats;
|
||||
printf("**********************\n");
|
||||
float g_avg_gen_time = 0;
|
||||
float g_avg_crossover_time = 0;
|
||||
float g_avg_mutate_time = 0;
|
||||
float g_avg_fitness_time = 0;
|
||||
float g_avg_sorting_time = 0;
|
||||
float g_avg_overhead_time = 0;
|
||||
float g_progress_per = 0;
|
||||
float g_best_fitness = strat.higher_fitness_is_better ? FLT_MIN : FLT_MAX;
|
||||
|
||||
complete = true;
|
||||
|
||||
for (int i = 0; i < stats.len; i++) {
|
||||
lock(stats[i].m);
|
||||
complete &= stats[i].done;
|
||||
|
||||
int end = stats[i].gen_time.end-1;
|
||||
|
||||
float gen_time = to_s(stats[i].gen_time[end]);
|
||||
float crossover_time = to_s(stats[i].crossover_time[end]);
|
||||
float mutate_time = to_s(stats[i].mutate_time[end]);
|
||||
float fitness_time = to_s(stats[i].fitness_time[end]);
|
||||
float sorting_time = to_s(stats[i].sorting_time[end]);
|
||||
float progress_per = static_cast<float>(stats[i].gen) / static_cast<float>(strat.num_generations) * 100;
|
||||
float best_score = back(stats[i].best_cell_fitness);
|
||||
|
||||
float overhead = max(0, gen_time - (crossover_time + mutate_time + fitness_time + sorting_time));
|
||||
|
||||
float overhead_per = overhead / gen_time * 100;
|
||||
|
||||
g_avg_gen_time += gen_time;
|
||||
g_avg_crossover_time += crossover_time;
|
||||
g_avg_mutate_time += mutate_time;
|
||||
g_avg_fitness_time += fitness_time;
|
||||
g_avg_sorting_time += sorting_time;
|
||||
g_progress_per += progress_per;
|
||||
g_best_fitness = better(strat.higher_fitness_is_better, best_score, g_best_fitness);
|
||||
|
||||
g_avg_overhead_time += overhead;
|
||||
|
||||
printf("%d, Progress %d/%d, Top: %.5e, Overhead Per: %.4f%%, Gen: %.4f, Overhead: %.4f, Cross: %.4f (s), Mutate: %.4f (s), Fitness: %.4f (s), Sorting: %.4f (s)\n", i, stats[i].gen, strat.num_generations, best_score, overhead_per, gen_time, overhead, crossover_time, mutate_time, fitness_time, sorting_time);
|
||||
unlock(stats[i].m);
|
||||
}
|
||||
|
||||
g_avg_gen_time /= stats.len;
|
||||
g_avg_crossover_time /= stats.len;
|
||||
g_avg_mutate_time /= stats.len;
|
||||
g_avg_fitness_time /= stats.len;
|
||||
g_avg_sorting_time /= stats.len;
|
||||
g_progress_per /= stats.len;
|
||||
|
||||
g_avg_overhead_time /= stats.len;
|
||||
|
||||
float g_avg_overhead_per = g_avg_overhead_time / g_avg_gen_time * 100;
|
||||
|
||||
printf("GLOBAL, Progress %.1f%%, Top: %.5e, Overhead Per: %.4f%%, Gen: %.4f, Overhead: %.4f, Cross: %.4f (s), Mutate: %.4f (s), Fitness: %.4f (s), Sorting: %.4f (s)\n", g_progress_per, g_best_fitness, g_avg_overhead_per, g_avg_gen_time, g_avg_overhead_time, g_avg_crossover_time, g_avg_mutate_time, g_avg_fitness_time, g_avg_sorting_time);
|
||||
|
||||
if (complete) break;
|
||||
}
|
||||
|
||||
for (int i = 0; i < threads.len; i++) {
|
||||
join(threads[i]);
|
||||
}
|
||||
|
||||
T best_cell;
|
||||
// TODO: bad
|
||||
float best_score = strat.higher_fitness_is_better ? FLT_MIN : FLT_MAX;
|
||||
for (int i = 0; i < stats.len; i++) {
|
||||
float score = back(stats[i].best_cell_fitness);
|
||||
if (strat.higher_fitness_is_better ? score > best_score : score < best_score) {
|
||||
best_cell = back(stats[i].best_cells);
|
||||
best_score = score;
|
||||
}
|
||||
}
|
||||
|
||||
return best_cell;
|
||||
}
|
||||
|
||||
template<class T> WorkQueue<T> make_work_queue(int len, int batch_size) {
|
||||
|
||||
Reference in New Issue
Block a user