├── chess_tuner ├── __init__.py ├── arena.py └── chess_tuner.py ├── static └── demo.png ├── .gitignore ├── demo2.py ├── test.py ├── README.md ├── nobo.py └── LICENSE /chess_tuner/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /static/demo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/thomasahle/noisy-bayesian-optimization/HEAD/static/demo.png -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | *.swp 2 | *.swo 3 | .DS_Store 4 | __pycache__/ 5 | demo 6 | demo.py 7 | tune.log 8 | tune_debugfile 9 | tune_out.pgn 10 | *.pyc 11 | -------------------------------------------------------------------------------- /demo2.py: -------------------------------------------------------------------------------- 1 | import nobo, skopt, random 2 | 3 | def f(xs): 4 | x, y = xs 5 | return int(random.random() < abs(x) + abs(y)) 6 | 7 | bo = nobo.Optimizer([skopt.utils.Real(-1, 1), skopt.utils.Real(-1, 1)]) 8 | 9 | for j in range(50): 10 | x = bo.ask(verbose=False) 11 | y = f(x) 12 | print(x, y) 13 | bo.tell(x, y) 14 | 15 | x, lo, y, hi = bo.get_best() 16 | print(f'Best: {x}, f(x) = {y:.3} +/- {(hi-lo)/2:.3}') 17 | bo.plot() 18 | -------------------------------------------------------------------------------- /test.py: -------------------------------------------------------------------------------- 1 | import unittest 2 | import nobo 3 | import random 4 | import skopt 5 | 6 | # TODO: Fix random state 7 | 8 | class TestOptimizer(unittest.TestCase): 9 | 10 | def _simple_opt(self, dims, fun, niter, maximize=False): 11 | bo = nobo.Optimizer(dims, maximize=maximize) 12 | for j in range(niter): 13 | x = bo.ask() 14 | bo.tell(x, fun(x)) 15 | return bo.get_best() 16 | 17 | def test_deterministic(self): 18 | def f(x): 19 | return int(abs(x[0]) > .1) 20 | 21 | x, lo, y, hi = self._simple_opt([skopt.utils.Real(-1, 1)], f, 30) 22 | self.assertAlmostEqual(x[0], 0, 1) 23 | self.assertAlmostEqual(y, 0, 1) 24 | self.assertLess(y, hi) 25 | self.assertGreater(y, lo) 26 | 27 | def test_noisy(self): 28 | def f(x): 29 | return int(random.random() < x[0]**2) 30 | 31 | x, lo, y, hi = self._simple_opt([skopt.utils.Real(-1, 1)], f, 100) 32 | self.assertAlmostEqual(x[0], 0, 1) 33 | self.assertAlmostEqual(y, 0, 1) 34 | self.assertLess(y, hi) 35 | self.assertGreater(y, lo) 36 | 37 | def test_very_noisy(self): 38 | def f(x, noise=0.5): 39 | if random.random() > noise: 40 | return int(random.random() < x[0]**2) 41 | return int(random.random() < .5) 42 | 43 | x, lo, y, hi = self._simple_opt([skopt.utils.Real(-1, 1)], f, 100) 44 | self.assertAlmostEqual(x[0], 0, 1) 45 | self.assertAlmostEqual(y, .25, 1) 46 | self.assertLess(y, hi) 47 | self.assertGreater(y, lo) 48 | 49 | def test_2d_maximization(self): 50 | def f(x): 51 | return int(random.random() > (abs(x[0])+abs(x[1]))/2) 52 | 53 | x, lo, y, hi = self._simple_opt( 54 | [skopt.utils.Real(-1, 1), skopt.utils.Real(-1, 1)], 55 | f, 100, maximize=True) 56 | self.assertAlmostEqual(x[0], 0, 1) 57 | self.assertAlmostEqual(x[1], 0, 1) 58 | self.assertAlmostEqual(y, 0, 1) 59 | self.assertLess(y, hi) 60 | self.assertGreater(y, lo) 61 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Nobo (or noisy-bayesian-optimization) 2 | Nobo is a library for optimizing very noisy functions using [bayesian optimization](https://en.wikipedia.org/wiki/Bayesian_optimization). 3 | In particular functions `p(x) -> [0,1]` for which you only observe a random variable `V_x ~ Bernoulli(p(x))`. 4 | This might model A/B testing or a game playing program. 5 | 6 | The following example shows how to use Nobo to find the minimum of `p(x) = x^2` over `[-1, 1]` in this situation: 7 | 8 | ```python 9 | >>> import nobo, skopt, random 10 | >>> 11 | >>> def f(x): 12 | >>> return int(random.random() < x[0]**2) 13 | >>> 14 | >>> bo = nobo.Optimizer([skopt.utils.Real(-1, 1)]) 15 | >>> for j in range(50): 16 | >>> x = bo.ask(verbose=True) 17 | >>> bo.tell(x, f(x)) 18 | >>> 19 | >>> x, lo, y, hi = bo.get_best() 20 | >>> print(f'Best: {x}, f(x) = {y:.3} +/- {(hi-lo)/2:.3}') 21 | >>> bo.plot() 22 | ... 23 | Best: [0.05464264314195355], f(x) = 0.018 +/- 0.0213 24 | ``` 25 | ![Screenshot](https://raw.githubusercontent.com/thomasahle/noisy-bayesian-optimization/master/static/demo.png) 26 | 27 | # Implementation 28 | Nobo uses [gpytorch](https://gpytorch.ai/) to do Gaussian Process Regression. This has the advantage over the classic [sciki-optimize](https://scikit-optimize.github.io/) `GaussianProcessRegressor` that [we can specify a non gaussian likelihood](https://gpytorch.readthedocs.io/en/latest/examples/04_Variational_and_Approximate_GPs/Non_Gaussian_Likelihoods.html) such as, in our case, Bernoulli. 29 | 30 | The acquisition function used is Lower Confidnce Bound, taking as the next `x` to investigate the `argmin_x mean(x) - log(n)*stddev(x)` where `n` is the number of points already considered. 31 | 32 | # Chess Tuner 33 | A classic use of Black Box Noisy Optimization is training the hyperparameters of game playing programs. 34 | In `chess/chess_tuner.py` we include a useful tool giving a complete playing and optimization pipeline. 35 | An example of usage is: 36 | 37 | ```bash 38 | $ python python -m chess_tuner.chess_tuner sunfish -n 1000 -movetime 40 -conrrency=20 39 | -book lines.pgn -games-file tune_out.pgn 40 | -opt eval_roughness 1 30 -opt qs_limit 1 400 41 | -log-file tune.log -debug tune_debugfile -conf engines.json -result-interval 10 42 | Loaded book with 173 positions 43 | Loading 20 engines 44 | Parsing options 45 | Reading tune.log 46 | Using [26.0, 72.0] => 0.0 from log-file 47 | Using [23.0, 72.0] => 0.0 from log-file 48 | Using [26.0, 140.0] => 0.0 from log-file 49 | Using [21.0, 175.0] => 1.0 from log-file 50 | ... 51 | Starting 2 games 113/1000 with {'eval_roughness': 5, 'qs_limit': 79} 52 | Starting 2 games 114/1000 with {'eval_roughness': 11, 'qs_limit': 115} 53 | Starting 2 games 115/1000 with {'eval_roughness': 27, 'qs_limit': 167} 54 | Starting 2 games 116/1000 with {'eval_roughness': 18, 'qs_limit': 176} 55 | Starting 2 games 117/1000 with {'eval_roughness': 13, 'qs_limit': 113} 56 | Starting 2 games 118/1000 with {'eval_roughness': 16, 'qs_limit': 17} 57 | ... 58 | Finished game 115 [27, 167] => 1.0 (1-0, 0-1) 59 | Starting 2 games 133/1000 with {'eval_roughness': 6, 'qs_limit': 194} 60 | Finished game 130 [16, 174] => 1.0 (1-0, 0-1) 61 | Starting 2 games 134/1000 with {'eval_roughness': 19, 'qs_limit': 176} 62 | Summarizing best values 63 | Best expectation (κ=0.0): [7, 78] = 0.363 ± 0.424 (ELO-diff 132.3 ± 164.0) 64 | ... 65 | ``` 66 | 67 | Here two (uci or xboard) parameters `eval_roughness` and `qs_limit` are optimized. 68 | Games are played against the unoptimized engine. 69 | For all available options see `python -m chess_tuner.chess_tuner --help`. 70 | The code is a fork of the [fastchess](https://github.com/thomasahle/fastchess) chess tuner, which used normal gaussian bayesian optimization, and thus would often converge to the wrong values. 71 | 72 | # Installation 73 | 74 | You need to `pip install numpy scipy scikit-optimize gpytorch torch matplotlib`. 75 | To run the chess tuner you also need `pip install chess`. 76 | You can then `git clone git@github.com:thomasahle/noisy-bayesian-optimization.git` and run `python noisy-bayesian-optimization/chess/chess_tuner.py` directly. 77 | -------------------------------------------------------------------------------- /chess_tuner/arena.py: -------------------------------------------------------------------------------- 1 | import asyncio 2 | import itertools 3 | import random 4 | import logging 5 | 6 | import chess.pgn 7 | import chess.engine 8 | import chess 9 | from chess import WHITE, BLACK 10 | 11 | 12 | class Arena: 13 | MATE_SCORE = 32767 14 | 15 | def __init__(self, enginea, engineb, book, limit, max_len, win_adj_count, win_adj_score): 16 | self.enginea = enginea 17 | self.engineb = engineb 18 | self.book = book 19 | self.limit = limit 20 | self.max_len = max_len 21 | self.win_adj_count = win_adj_count 22 | self.win_adj_score = win_adj_score 23 | 24 | def adjudicate(self, score_hist): 25 | if len(score_hist) > self.max_len: 26 | return '1/2-1/2' 27 | # Note win_adj_count is in moves, not plies 28 | count_max = self.win_adj_count * 2 29 | if count_max > len(score_hist): 30 | return None 31 | # Test if white has been winning. Notice score_hist is from whites pov. 32 | if all(v >= self.win_adj_score for v in score_hist[-count_max:]): 33 | return '1-0' 34 | # Test if black has been winning 35 | if all(v <= -self.win_adj_score for v in score_hist[-count_max:]): 36 | return '0-1' 37 | return None 38 | 39 | async def play_game(self, init_node, game_id, white, black): 40 | """ Yields (play, error) tuples. Also updates the game with headers and moves. """ 41 | try: 42 | game = init_node.root() 43 | node = init_node 44 | score_hist = [] 45 | for ply in itertools.count(int(node.board().turn == BLACK)): 46 | board = node.board() 47 | 48 | adj_result = self.adjudicate(score_hist) 49 | if adj_result is not None: 50 | game.headers.update({ 51 | 'Result': adj_result, 52 | 'Termination': 'adjudication' 53 | }) 54 | return 55 | 56 | if board.is_game_over(claim_draw=True): 57 | result = board.result(claim_draw=True) 58 | game.headers["Result"] = result 59 | return 60 | 61 | # Try to actually make a move 62 | play = await (white, black)[ply % 2].play( 63 | board, self.limit, game=game_id, 64 | info=chess.engine.INFO_BASIC | chess.engine.INFO_SCORE) 65 | yield play, None 66 | 67 | if play.resigned: 68 | game.headers.update({'Result': ('0-1', '1-0')[ply % 2]}) 69 | node.comment += f'; {("White","Black")[ply%2]} resgined' 70 | return 71 | 72 | node = node.add_variation(play.move, comment= 73 | f'{play.info.get("score",0)}/{play.info.get("depth",0)}' 74 | f' {play.info.get("time",0)}s') 75 | 76 | # Adjudicate game by score, save score in wpov 77 | try: 78 | score_hist.append(play.info['score'].white().score( 79 | mate_score=max(self.win_adj_score, Arena.MATE_SCORE))) 80 | except KeyError: 81 | logging.debug('Engine didn\'t return a score for adjudication. Assuming 0.') 82 | score_hist.append(0) 83 | 84 | except (asyncio.CancelledError, KeyboardInterrupt) as e: 85 | print('play_game Cancelled') 86 | # We should get CancelledError when the user pressed Ctrl+C 87 | game.headers.update({'Result': '*', 'Termination': 'unterminated'}) 88 | node.comment += '; Game was cancelled' 89 | await asyncio.wait([white.quit(), black.quit()]) 90 | yield None, e 91 | except chess.engine.EngineError as e: 92 | game.headers.update( 93 | {'Result': ('0-1', '1-0')[ply % 2], 'Termination': 'error'}) 94 | node.comment += f'; {("White","Black")[ply%2]} terminated: {e}' 95 | yield None, e 96 | 97 | async def configure(self, args): 98 | # We configure enginea, engineb is our unchanged opponent. 99 | # Maybe this should be refactored. 100 | self.enginea.id['args'] = args 101 | self.engineb.id['args'] = {} 102 | try: 103 | await self.enginea.configure(args) 104 | except chess.engine.EngineError as e: 105 | print(f'Unable to start game {e}') 106 | return [], 0 107 | 108 | async def run_games(self, game_id=0, games_played=2): 109 | score = 0 110 | games = [] 111 | assert games_played % 2 == 0 112 | for r in range(games_played//2): 113 | init_board = random.choice(self.book) 114 | for color in [WHITE, BLACK]: 115 | white, black = (self.enginea, self.engineb) if color == WHITE \ 116 | else (self.engineb, self.enginea) 117 | game_round = games_played * game_id + color + 2*r 118 | game = chess.pgn.Game({ 119 | 'Event': 'Tune.py', 120 | 'White': white.id['name'], 121 | 'WhiteArgs': repr(white.id['args']), 122 | 'Black': black.id['name'], 123 | 'BlackArgs': repr(black.id['args']), 124 | 'Round': game_round 125 | }) 126 | games.append(game) 127 | # Add book moves 128 | game.setup(init_board.root()) 129 | node = game 130 | for move in init_board.move_stack: 131 | node = node.add_variation(move, comment='book') 132 | # Run engines 133 | async for _play, er in self.play_game(node, game_round, white, black): 134 | # If an error occoured, return as much as we got 135 | if er is not None: 136 | return games, score, er 137 | result = game.headers["Result"] 138 | if result == '1-0' and color == WHITE or result == '0-1' and color == BLACK: 139 | score += 1 140 | if result == '1-0' and color == BLACK or result == '0-1' and color == WHITE: 141 | score -= 1 142 | return games, score/games_played, None 143 | 144 | 145 | class ArenaRunner: 146 | def __init__(self, engines, opt, x_to_args, n_games, concurrency=1, games_per_encounter=1): 147 | self.engines = engines 148 | self.opt = opt 149 | self.concurrency = concurrency 150 | assert len(engines) == concurrency 151 | self.started = 0 152 | self.games_per_encounter = games_per_encounter 153 | self.x_to_args = x_to_args 154 | self.n_games = n_games 155 | 156 | def _on_done(self, task): 157 | if task.exception(): 158 | logging.error('Error while excecuting game') 159 | task.print_stack() 160 | 161 | def _new_game(self, arena): 162 | async def routine(game_id): 163 | x = await self.opt.ask() 164 | engine_args = self.x_to_args(x) 165 | print(f'Starting {self.games_per_encounter} games {game_id}/{self.n_games} with {engine_args}') 166 | await arena.configure(engine_args) 167 | res = await arena.run_games(game_id=game_id, games_played=self.games_per_encounter) 168 | return x, res 169 | task = asyncio.create_task(routine(self.started)) 170 | # We tag the task with some attributes that we need when it finishes. 171 | setattr(task, 'tune_arena', arena) 172 | setattr(task, 'tune_game_id', self.started) 173 | task.add_done_callback(self._on_done) 174 | self.started += 1 175 | return task 176 | 177 | async def run(self, arena_args): 178 | # Find how many games are already in the optimizer 179 | self.started = await self.opt.size() 180 | 181 | tasks = [] 182 | if self.n_games - self.started > 0: 183 | xs = await self.opt.ask(min(self.concurrency, self.n_games - self.started)) 184 | else: 185 | xs = [] 186 | assert len(xs) <= self.concurrency 187 | 188 | for conc_id, x_init in enumerate(xs): 189 | enginea, engineb = self.engines[conc_id] 190 | arena = Arena(enginea, engineb, *arena_args) 191 | tasks.append(self._new_game(arena)) 192 | 193 | while tasks: 194 | done, pending = await asyncio.wait(tasks, return_when=asyncio.FIRST_COMPLETED) 195 | tasks = list(pending) 196 | for task in done: 197 | arena, game_id = task.tune_arena, task.tune_game_id 198 | x, (games, y, er) = task.result() 199 | 200 | yield (game_id, x, y, games, er) 201 | 202 | await self.opt.tell(x, (y+1)/2) # Format in [0,1] 203 | 204 | if self.started < self.n_games: 205 | tasks.append(self._new_game(arena)) 206 | 207 | -------------------------------------------------------------------------------- /nobo.py: -------------------------------------------------------------------------------- 1 | import os 2 | from gpytorch.variational import UnwhitenedVariationalStrategy, VariationalStrategy 3 | from gpytorch.variational import CholeskyVariationalDistribution 4 | from gpytorch.models import ApproximateGP 5 | import torch 6 | import gpytorch 7 | import numpy as np 8 | from scipy.stats import norm as normal 9 | import skopt.utils 10 | 11 | from botorch.models.gpytorch import GPyTorchModel 12 | from botorch.acquisition import UpperConfidenceBound, qNoisyExpectedImprovement 13 | from botorch.optim import optimize_acqf 14 | from botorch.fit import fit_gpytorch_model 15 | #class _GPModel(ApproximateGP): 16 | class _GPModel(ApproximateGP, GPyTorchModel): 17 | _num_outputs = 1 18 | 19 | def __init__(self, train_x, sigma=1): 20 | variational_distribution = CholeskyVariationalDistribution( 21 | train_x.size(0)) 22 | variational_strategy = VariationalStrategy( 23 | self, train_x, variational_distribution, learn_inducing_locations=False) 24 | super(_GPModel, self).__init__(variational_strategy) 25 | self.mean_module = gpytorch.means.ConstantMean() 26 | self.covar_module = gpytorch.kernels.ScaleKernel( 27 | #gpytorch.kernels.RBFKernel(lengthscale_prior=gpytorch.priors.NormalPrior(0, sigma))) 28 | gpytorch.kernels.RBFKernel()) 29 | 30 | def forward(self, x): 31 | mean_x = self.mean_module(x) 32 | covar_x = self.covar_module(x) 33 | latent_pred = gpytorch.distributions.MultivariateNormal( 34 | mean_x, covar_x) 35 | return latent_pred 36 | 37 | 38 | class Optimizer: 39 | def __init__(self, dimensions, initial_points=10, maximize=False): 40 | self.space = skopt.utils.Space(dimensions) 41 | self.d = len(dimensions) 42 | self.initial_points = initial_points 43 | self.likelihood = gpytorch.likelihoods.BernoulliLikelihood() 44 | self.maximize = maximize 45 | self.xs = [] 46 | self.ys = [] 47 | 48 | def _train(self, iterations, lr, verbose=False): 49 | train_x = torch.cat(self.xs, dim=0).float() 50 | train_y = torch.tensor(self.ys, dtype=torch.float).float() 51 | model = _GPModel(train_x, sigma=1) 52 | optimizer = torch.optim.Adam(model.parameters(), lr=lr) 53 | mll = gpytorch.mlls.VariationalELBO( 54 | self.likelihood, model, train_y.numel()) 55 | 56 | # Find optimal model hyperparameters 57 | model.train() 58 | self.likelihood.train() 59 | #losses = [] 60 | for i in range(iterations): 61 | optimizer.zero_grad() 62 | try: 63 | output = model(train_x) 64 | except RuntimeError: 65 | if lr > 1e-6: 66 | print('Error in training. Trying again with lower lr') 67 | return self._train(iterations, lr/2, verbose) 68 | else: 69 | print(f'Stopping after {i} iterations. Lr = {lr}') 70 | break 71 | loss = -mll(output, train_y) 72 | loss.backward() 73 | #losses.append(loss.item()) 74 | if verbose: 75 | print('Iter %d/%d - Loss: %.3f' % 76 | (i+1, iterations, loss.item())) 77 | #if len(losses) >= iterations and np.argmin(losses)/len(losses) < 3/4: 78 | #break 79 | optimizer.step() 80 | 81 | model.eval() 82 | self.likelihood.eval() 83 | return model 84 | 85 | def _random_in_bounds(self, n=1): 86 | return self.space.transform(self.space.rvs(n)) 87 | 88 | def ask2(self, n=1, iterations=10, lr=0.1, verbose=False): 89 | if n > 1: 90 | xs = list(self._random_in_bounds(n-1)) 91 | xs += self.ask(1, iterations, lr, verbose) 92 | return xs 93 | # Use random points in the beginning 94 | if len(self.xs) < self.initial_points: 95 | return self._random_in_bounds()[0] 96 | # Train new model 97 | kappa = np.sqrt(2*np.log(len(self.xs)+1)) 98 | model = self._train(iterations, lr, verbose=verbose) 99 | 100 | #UCB = UpperConfidenceBound(model, beta=kappa, maximize=self.maximize) 101 | UCB = UpperConfidenceBound(model, beta=.1, maximize=self.maximize) 102 | print(UCB) 103 | print(UCB.maximize) 104 | 105 | #test_x = torch.linspace(-1, 1, 50) 106 | #post = model.posterior(test_x) 107 | 108 | ucb_cands, ucb_vals = optimize_acqf( 109 | acq_function=UCB, 110 | bounds=torch.tensor(self.space.transformed_bounds).T.float(), 111 | q=1, 112 | num_restarts=1, 113 | raw_samples=25, 114 | ) 115 | print(ucb_cands, ucb_vals) 116 | return ucb_cands 117 | 118 | def ask3(self, n=1, iterations=10, lr=0.1, verbose=False): 119 | # Use random points in the beginning 120 | if len(self.xs) < self.initial_points: 121 | return self._random_in_bounds()[0] 122 | 123 | train_x = torch.cat(self.xs, dim=0).float() 124 | #X_baseline = train_x.unsqueeze(-1) # in botorch the feature dimension is assumed explicit 125 | X_baseline = train_x 126 | model = self._train(iterations, lr, verbose=verbose) 127 | 128 | qNEI = qNoisyExpectedImprovement(model, X_baseline=X_baseline, prune_baseline=True, 129 | #maximize=self.maximize 130 | ) 131 | 132 | qnei_cands, qnei_vals = optimize_acqf( 133 | acq_function=qNEI, 134 | bounds=torch.tensor(self.space.transformed_bounds).T.float(), 135 | q=1, 136 | num_restarts=1, 137 | raw_samples=25, 138 | ) 139 | 140 | print(qnei_cands, qnei_vals) 141 | return qnei_cands 142 | 143 | def ask(self, n=None, iterations=10, lr=0.1, verbose=False): 144 | # If n is not given, we return a single point. 145 | # If n is given, including if n=1, we return a list of that many points 146 | if n is None: 147 | x = self.ask(1, iterations, lr, verbose) 148 | return x[0] 149 | 150 | # TODO: Find multiple points more intelligently. 151 | # Can also use yield to allow called to get started early. 152 | if n > 1: 153 | return list(self._random_in_bounds(n-1)) \ 154 | + self.ask(1, iterations, lr, verbose) 155 | 156 | # Use random points in the beginning 157 | if len(self.xs) < self.initial_points: 158 | return self._random_in_bounds() 159 | 160 | # Train new model 161 | kappa = np.sqrt(2*np.log(len(self.xs)+1)) 162 | #kappa = np.log(len(self.xs)+1) 163 | model = self._train(iterations, lr, verbose=verbose) 164 | with torch.no_grad(): 165 | # TODO: We should do this instead using optimization, or using 166 | # the training loop (above) directly 167 | test_x = self._random_in_bounds(n=100) 168 | try: 169 | latent_pred = model(torch.tensor(test_x).float()) 170 | except RuntimeError: 171 | print('Error in running model. Returning random point.') 172 | return self._random_in_bounds() 173 | # Using Lower Confidnce Bound. Could be changed to something else 174 | # like expected improvement. 175 | xstar = test_x[np.argmin( 176 | latent_pred.mean - kappa*latent_pred.stddev)] 177 | return self.space.inverse_transform(xstar.reshape(1,-1)) 178 | 179 | def tell(self, x, y): 180 | #print(f'tell({x}, {y})') 181 | assert 0-1e-5 <= y <= 1+1e-5, f'y ({y}) not in range [0, 1]' 182 | x = self.space.transform([x]) 183 | self.xs.append(torch.tensor(x).float()) 184 | if self.maximize: 185 | y = 1-y 186 | self.ys.append(float(y)) 187 | 188 | def get_best(self, iterations=50, lr=0.1, kappa=0, restarts=0): 189 | best = 1 190 | for j in range(restarts+1): 191 | model = self._train(iterations, lr) 192 | with torch.no_grad(): 193 | # TODO: We should do this instead using optimization, or using 194 | # the training loop (above) directly 195 | test_x = self._random_in_bounds(n=500) 196 | try: 197 | latent_pred = model(torch.tensor(test_x).float()) 198 | except RuntimeError: 199 | print('Error in running model. Returning 0') 200 | return 0, 0, 0, 0 201 | # Using Lower Confidnce Bound. Could be changed to something else 202 | # like expected improvement. 203 | vals = latent_pred.mean - kappa*latent_pred.stddev 204 | i = np.argmin(vals) 205 | mean = normal.cdf(vals[i]) 206 | if mean < best: 207 | best = mean 208 | x = self.space.inverse_transform(test_x[i].reshape(1,-1))[0] 209 | lower = normal.cdf(latent_pred.mean[i] - latent_pred.stddev[i]) 210 | upper = normal.cdf(latent_pred.mean[i] + latent_pred.stddev[i]) 211 | if self.maximize: 212 | return x, 1-upper, 1-mean, 1-lower 213 | return x, lower, mean, upper 214 | 215 | def size(self): 216 | return len(self.xs) 217 | 218 | def plot(self, iterations=50, lr=0.1): 219 | assert self.d <= 2 220 | 221 | model = self._train(iterations, lr) 222 | 223 | if self.d == 1: 224 | self._plot_1d(model) 225 | if self.d == 2: 226 | self._plot_2d(model) 227 | 228 | def _plot_1d(self, model): 229 | from matplotlib import pyplot as plt 230 | import matplotlib 231 | matplotlib.use('tkagg') 232 | 233 | with torch.no_grad(): 234 | test_x = self._random_in_bounds(n=100) 235 | test_x = np.sort(test_x, axis=0) 236 | latent_pred = model(torch.tensor(test_x).float()) 237 | 238 | kappa = np.log(len(self.xs)+1) 239 | _fig, ax = plt.subplots(1, 1, figsize=(4, 3)) 240 | ax.plot(self.xs, self.ys, 'k*', label='Observed Data') 241 | test_x = test_x.reshape(-1) 242 | mean = normal.cdf(latent_pred.mean.numpy()) 243 | lower = normal.cdf( 244 | (latent_pred.mean - kappa*latent_pred.stddev).numpy()) 245 | upper = normal.cdf( 246 | (latent_pred.mean + kappa*latent_pred.stddev).numpy()) 247 | ax.plot(test_x, mean, label='Mean') 248 | ax.fill_between(test_x, upper, lower, color="#b9cfe7", edgecolor="") 249 | ax.set_ylim([-1, 2]) 250 | ax.legend() 251 | plt.show() 252 | 253 | def _plot_2d(self, model): 254 | from matplotlib import pyplot as plt 255 | import matplotlib 256 | matplotlib.use('tkagg') 257 | 258 | with torch.no_grad(): 259 | x = self._random_in_bounds(n=1000) 260 | latent_pred = model(torch.tensor(x).float()) 261 | mean = normal.cdf(latent_pred.mean.numpy()) 262 | 263 | fig, ax = plt.subplots() 264 | cs = ax.tricontourf(x[:,0], x[:,1], mean) 265 | x = torch.cat(self.xs, dim=0).numpy().reshape(-1,2) 266 | ax.plot(x[:,0], x[:,1], 'ko', ms=3) 267 | cbar = fig.colorbar(cs) 268 | plt.show() 269 | 270 | 271 | import multiprocessing as mp 272 | import concurrent, asyncio 273 | class RPC(mp.Process): 274 | def __init__(self, cls, *args, **kwargs): 275 | super(RPC, self).__init__() 276 | self.obj = cls(*args, **kwargs) 277 | self.in_queue = mp.Queue() 278 | self.out_queue = mp.Queue() 279 | self.ex = concurrent.futures.ThreadPoolExecutor(max_workers=1) 280 | def _submit(self, f, *args, **kwargs): 281 | return asyncio.wrap_future(self.ex.submit(f, *args, **kwargs)) 282 | def __getattr__(self, attr): 283 | async def inner(*args, **kwargs): 284 | recv_conn, send_conn = mp.Pipe() 285 | await self._submit(self.in_queue.put, (attr, send_conn, args, kwargs)) 286 | res = await self._submit(recv_conn.recv) 287 | recv_conn.close() 288 | send_conn.close() 289 | return res 290 | return inner 291 | def run(self): 292 | while True: 293 | attr, conn, args, kwargs = self.in_queue.get() 294 | method = getattr(self.obj, attr) 295 | try: 296 | result = method(*args, **kwargs) 297 | except BaseException as err: 298 | conn.send(err) 299 | else: 300 | conn.send(result) 301 | def close(self): 302 | raise NotImplemented 303 | 304 | -------------------------------------------------------------------------------- /chess_tuner/chess_tuner.py: -------------------------------------------------------------------------------- 1 | import math 2 | import concurrent 3 | import hashlib 4 | import functools 5 | import sys 6 | import json 7 | import pathlib 8 | import asyncio 9 | import argparse 10 | import textwrap 11 | import warnings 12 | import itertools 13 | import re 14 | import os 15 | import logging 16 | from collections import namedtuple 17 | 18 | import chess.pgn 19 | import chess.engine 20 | import chess 21 | import numpy as np 22 | import nobo 23 | import skopt 24 | 25 | from .arena import Arena, ArenaRunner 26 | 27 | warnings.filterwarnings( 28 | 'ignore', 29 | message='The objective has been evaluated at this point before.') 30 | 31 | class Formatter(argparse.HelpFormatter): 32 | 33 | def _fill_text(self, text, width, indent): 34 | return ''.join(indent + line for line in text.splitlines(keepends=True)) 35 | 36 | def _get_help_string(self, action): 37 | help = action.help 38 | if not action.default: 39 | return help 40 | if '%(default)' not in action.help: 41 | if action.default is not argparse.SUPPRESS: 42 | defaulting_nargs = [argparse.OPTIONAL, argparse.ZERO_OR_MORE] 43 | if action.option_strings or action.nargs in defaulting_nargs: 44 | help += ' (default: %(default)s)' 45 | return help 46 | 47 | parser = argparse.ArgumentParser( 48 | formatter_class=Formatter, 49 | fromfile_prefix_chars='@', 50 | usage='%(prog)s ENGINE_NAME [options]', 51 | description=textwrap.dedent(''' 52 | Tune.py is a tool that allows you to tune chess engines with black 53 | box optimization through Scikit-Optimize (or skopt). Engine 54 | communication is handled through python-chess, so all you need is an 55 | uci or cecp compatible engine supporting options, and you are set! 56 | \n\n 57 | Simple example for tuning the MilliCpuct option of fastchess: 58 | $ python tune.py fastchess -opt MilliCpuct 59 | \n\n 60 | Tune.py uses an engine.json file to load engines. A simple such file 61 | is provided in the git repositiory and looks something like this: 62 | [{ 63 | "name": "stockfish", 64 | "command": "stockfish", 65 | "protocol": "uci" 66 | }] 67 | \n\n 68 | Tip: Use `-log-file data.log` to save the results so you can easily 69 | recover from a crash, or try new arguments. 70 | \n 71 | Tip: If you have too many options to handle, tune.py can read its 72 | arguemnts from a file with `tune.py @argumentfile`. 73 | '''), 74 | ) 75 | parser.add_argument('-debug', nargs='?', metavar='PATH', const=sys.stdout, 76 | default=None, type=pathlib.Path, 77 | help='Enable debugging of engines.') 78 | parser.add_argument('-log-file', metavar='PATH', type=pathlib.Path, 79 | help='Used to recover from crashes') 80 | parser.add_argument('-n', type=int, default=100, 81 | help='Number of iterations') 82 | parser.add_argument('-concurrency', type=int, default=1, metavar='N', 83 | help='Number of concurrent games') 84 | parser.add_argument('-games-file', metavar='PATH', type=pathlib.Path, 85 | help='Store all games to this pgn') 86 | parser.add_argument('-result-interval', metavar='N', type=int, default=50, 87 | help='How often to print the best estimate so far. 0 for never') 88 | 89 | engine_group = parser.add_argument_group('Engine options') 90 | engine_group.add_argument('engine', metavar='ENGINE_NAME', 91 | help='Engine to tune') 92 | engine_group.add_argument('-conf', type=pathlib.Path, metavar='PATH', 93 | help='Engines.json file to load from') 94 | engine_group.add_argument('-opp-engine', metavar='ENGINE_NAME', 95 | help='Tune against a different engine') 96 | 97 | games_group = parser.add_argument_group('Games format') 98 | games_group.add_argument('-book', type=pathlib.Path, metavar='PATH', 99 | help='pgn file with opening lines.') 100 | games_group.add_argument('-n-book', type=int, default=10, metavar='N', 101 | help='Length of opening lines to use in plies.') 102 | games_group.add_argument('-games-per-encounter', type=int, default=2, metavar='N', 103 | help='Number of book positions to play at each set of argument explored.') 104 | games_group.add_argument('-max-len', type=int, default=10000, metavar='N', 105 | help='Maximum length of game in plies before termination.') 106 | games_group.add_argument('-win-adj', nargs='*', metavar='ADJ', 107 | help='Adjudicate won game. Usage: ' 108 | '-win-adj count=4 score=400 ' 109 | 'If the last 4 successive moves of white had a score of ' 110 | '400 cp or more and the last 4 successive moves of black ' 111 | 'had a score of -400 or less then that game will be ' 112 | 'adjudicated to a win for white. When the situation is ' 113 | 'reversed black would win. ' 114 | f'Default values: count=4, score={Arena.MATE_SCORE}') 115 | subgroup = games_group.add_mutually_exclusive_group(required=True) 116 | subgroup.add_argument('-movetime', type=int, metavar='MS', 117 | help='Time per move in ms') 118 | subgroup.add_argument('-nodes', type=int, metavar='N', 119 | help='Nodes per move') 120 | 121 | tune_group = parser.add_argument_group('Options to tune') 122 | tune_group.add_argument('-opt', nargs='+', action='append', default=[], 123 | metavar=('NAME', 'LOWER, UPPER'), 124 | help='Integer option to tune.') 125 | tune_group.add_argument('-c-opt', nargs='+', action='append', default=[], 126 | metavar=('NAME', 'VALUE'), 127 | help='Categorical option to tune') 128 | 129 | group = parser.add_argument_group('Optimization parameters') 130 | group.add_argument('-base-estimator', default='GP', metavar='EST', 131 | help='One of "GP", "RF", "ET", "GBRT"') 132 | group.add_argument('-n-initial-points', type=int, default=10, metavar='N', 133 | help='Number of points chosen before approximating with base estimator.') 134 | group.add_argument('-acq-func', default='gp_hedge', metavar='FUNC', 135 | help='Can be either of "LCB" for lower confidence bound.' 136 | ' "EI" for negative expected improvement.' 137 | ' "PI" for negative probability of improvement.' 138 | ' "gp_hedge" Probabilistically chooses one of the above' 139 | ' three acquisition functions at every iteration.') 140 | group.add_argument('-acq-optimizer', default='sampling', metavar='OPT', 141 | help='Either "sampling", "lbfgs" or "auto"') 142 | group.add_argument('-acq-n-points', default=10000, metavar='N', 143 | help='Number of points to sample when acq-optimizer = sampling.') 144 | group.add_argument('-acq-noise', default=10, metavar='VAR', 145 | help='For the Gaussian Process optimizer, use this to specify the' 146 | ' variance of the assumed noise. Larger values mean more exploration.') 147 | group.add_argument('-acq-xi', default=0.01, metavar='XI', type=float, 148 | help='Controls how much improvement one wants over the previous best' 149 | ' values. Used when the acquisition is either "EI" or "PI".') 150 | group.add_argument('-acq-kappa', default=1.96, metavar='KAPPA', type=float, 151 | help='Controls how much of the variance in the predicted values should be' 152 | ' taken into account. If set to be very high, then we are favouring' 153 | ' exploration over exploitation and vice versa. Used when the acquisition' 154 | ' is "LCB".') 155 | 156 | 157 | async def load_engine(engine_args, name, debug=False): 158 | assert engine_args and any(a['name'] == name for a in engine_args), \ 159 | f'Engine "{name}" was not found in engines.json file' 160 | args = next(a for a in engine_args if a['name'] == name) 161 | curdir = str(pathlib.Path(__file__).parent.parent) 162 | popen_args = {'env': {'PATH': os.environ['PATH']}} 163 | # Using $FILE in the workingDirectory allows an easy way to have engine.json 164 | # relative paths. 165 | args['command'] = args['command'].replace('$FILE', curdir) 166 | if 'workingDirectory' in args: 167 | popen_args['cwd'] = args['workingDirectory'].replace('$FILE', curdir) 168 | # Note: We don't currently support shell in the command. 169 | # We could do that using shutils. 170 | cmd = args['command'].split() 171 | # Shortcut for python engines who want to use the same ececutable as tune 172 | if cmd[0] == '$PYTHON': 173 | cmd[0] = sys.executable 174 | # Hack for Windows systems that don't understand popen(cwd=..) for some reason 175 | if os.name == 'nt': 176 | wd_cmd = pathlib.Path(popen_args['cwd'], cmd[0]) 177 | if wd_cmd.is_file(): 178 | cmd[0] = str(wd_cmd) 179 | if args['protocol'] == 'uci': 180 | _, engine = await chess.engine.popen_uci(cmd, **popen_args) 181 | elif args['protocol'] == 'xboard': 182 | _, engine = await chess.engine.popen_xboard(cmd, **popen_args) 183 | if hasattr(engine, 'debug'): 184 | engine.debug(debug) 185 | return engine 186 | 187 | 188 | def load_conf(conf): 189 | if not conf: 190 | path = pathlib.Path(__file__).parent.parent / 'engines.json' 191 | assert path.is_file(), 'No engines conf specified and unable to locate' \ 192 | ' engines.json file automatically.' 193 | return json.load(path.open()) 194 | else: 195 | assert conf.is_file(), f'Unable to open "{conf}"' 196 | return json.load(conf.open()) 197 | 198 | 199 | class DataLogger: 200 | def __init__(self, path, key): 201 | self.path = path 202 | self.key = key 203 | self.append_file = None 204 | 205 | def load(self): 206 | if not self.path.is_file(): 207 | print(f'Unable to open {self.path}') 208 | return 209 | print(f'Reading {self.path}') 210 | with self.path.open('r') as file: 211 | for line in file: 212 | obj = json.loads(line) 213 | if obj.get('args') == self.key: 214 | x, y = obj['x'], obj['y'] 215 | try: 216 | yield (x, y) 217 | except ValueError as e: 218 | print('Ignoring bad data point', e) 219 | 220 | def store(self, x, y): 221 | if not self.append_file: 222 | self.append_file = self.path.open('a') 223 | x = [xi if type(xi) == str else float(xi) for xi in x] 224 | y = float(y) 225 | print(json.dumps({'args': self.key, 'x': x, 'y': y}), 226 | file=self.append_file, flush=True) 227 | 228 | 229 | def load_book(path, n_book): 230 | if not path.is_file(): 231 | print(f'Error: Can\'t open book {path}.') 232 | return 233 | with open(path, encoding='latin-1') as file: 234 | for game in iter((lambda: chess.pgn.read_game(file)), None): 235 | board = game.board() 236 | for _, move in zip(range(n_book), game.mainline_moves()): 237 | board.push(move) 238 | yield board 239 | 240 | 241 | def parse_options(opts, copts, engine_options): 242 | dim_names = [] 243 | dimensions = [] 244 | for name, *lower_upper in opts: 245 | opt = engine_options.get(name) 246 | if not opt: 247 | if not lower_upper: 248 | print(f'Error: engine has no option {name}. For hidden options' 249 | ' you must specify lower and upper bounds.') 250 | continue 251 | else: 252 | print(f'Warning: engine has no option {name}') 253 | dim_names.append(name) 254 | lower, upper = map(int, lower_upper) if lower_upper else (opt.min, opt.max) 255 | dimensions.append(skopt.utils.Integer(lower, upper, name=name)) 256 | for name, *categories in copts: 257 | opt = engine_options.get(name) 258 | if not opt: 259 | if not categories: 260 | print(f'Error: engine has no option {name}. For hidden options' 261 | ' you must manually specify possible values.') 262 | continue 263 | else: 264 | print(f'Warning: engine has no option {name}') 265 | dim_names.append(name) 266 | cats = categories or opt.var 267 | cats = [opt.var.index(cat) for cat in cats] 268 | dimensions.append(skopt.utils.Categorical(cats, name=name)) 269 | if not dimensions: 270 | print('Warning: No options specified for tuning.') 271 | return dim_names, dimensions 272 | 273 | 274 | async def summarize(opt, steps, restarts=0): 275 | print('Summarizing best values') 276 | for kappa in [0] + list(np.logspace(-1, 1, steps-1)): 277 | x, lo, y, hi = await opt.get_best(kappa=kappa, restarts=restarts) 278 | # Optimizer uses [0,1]. We use [-1,1]. 279 | lo, y, hi = lo*2-1, y*2-1, hi*2-1 280 | def score_to_elo(score): 281 | if score <= -1: 282 | return float('inf') 283 | if score >= 1: 284 | return -float('inf') 285 | # Formula frorm https://www.chessprogramming.org/Match_Statistics 286 | return 400 * math.log10((1+score)/(1-score)) 287 | elo = score_to_elo(y) 288 | raw_pm = max(y-lo, hi-y) 289 | pm = max(abs(score_to_elo(hi) - elo), 290 | abs(score_to_elo(lo) - elo)) 291 | print(f'Best expectation (κ={kappa:.1f}): {x}' 292 | f' = {y:.3f} ± {raw_pm:.3f}' 293 | f' (ELO-diff {elo:.1f} ± {pm:.1f})') 294 | 295 | async def main(): 296 | args = parser.parse_args() 297 | 298 | if args.debug: 299 | if args.debug == sys.stdout: 300 | logging.basicConfig(level=logging.DEBUG) 301 | else: 302 | logging.basicConfig(level=logging.DEBUG, filename=args.debug, filemode='w') 303 | 304 | # Do not run the tuner if something is wrong with the adjudication option 305 | # that is set by the user. These options could be critical in tuning. 306 | win_adj_count, win_adj_score = 4, Arena.MATE_SCORE 307 | if args.win_adj: 308 | for n in args.win_adj: 309 | m = re.match('count=(\d+)', n) 310 | if m: 311 | win_adj_count = int(m.group(1)) 312 | m = re.match('score=(\d+)', n) 313 | if m: 314 | win_adj_score = int(m.group(1)) 315 | 316 | limit = chess.engine.Limit( 317 | nodes=args.nodes, 318 | time=args.movetime and args.movetime / 1000) 319 | 320 | assert args.games_per_encounter >= 2 and args.games_per_encounter % 2 == 0, \ 321 | 'Games per encounter must be even and >= 2.' 322 | 323 | if args.n > 1000: 324 | print(f'Running large number of games ({args.n} > 1000).') 325 | print('Bayesian Optimization may be slow for that many points. Consider increasing -games-per-encounter instead.') 326 | 327 | # Load book 328 | book = [] 329 | if args.book: 330 | book.extend(load_book(args.book, args.n_book)) 331 | print(f'Loaded book with {len(book)} positions') 332 | if not book: 333 | book.append(chess.Board()) 334 | print('No book. Starting every game from initial position.') 335 | 336 | # Load a pair of engines for each concurrency 337 | print(f'Loading {2*args.concurrency} engines') 338 | conf = load_conf(args.conf) 339 | engines = await asyncio.gather(*(asyncio.gather( 340 | load_engine(conf, args.engine), 341 | load_engine(conf, args.opp_engine or args.engine)) 342 | for _ in range(args.concurrency))) 343 | options = engines[0][0].options 344 | 345 | # Load options to tune 346 | print('Parsing options') 347 | dim_names, dimensions = parse_options(args.opt, args.c_opt, options) 348 | def x_to_args(x): 349 | args = {name: val for name, val in zip(dim_names, x)} 350 | for name, val in args.items(): 351 | option = options[name] 352 | if option.type == 'combo': 353 | args[name] = option.var[val] 354 | return args 355 | 356 | # Start optimizer process 357 | opt = nobo.RPC(nobo.Optimizer, dimensions, maximize=True) 358 | opt.start() 359 | 360 | # Load stored data 361 | if args.log_file: 362 | key_args = {} 363 | # Not all arguments change the result, so no need to keep them in the key. 364 | for arg_group in (games_group, tune_group, engine_group): 365 | for arg in arg_group._group_actions: 366 | key = arg.dest 367 | key_args[key] = getattr(args, key) 368 | key = repr(sorted(key_args.items())).encode() 369 | data_logger = DataLogger(args.log_file, key=hashlib.sha256(key).hexdigest()) 370 | for x, y in data_logger.load(): 371 | print(f'Using {x} => {y} from log-file') 372 | await opt.tell(x, (y+1)/2) # Format in [0,1] 373 | else: 374 | print('No -log-file set. Results won\'t be saved.') 375 | data_logger = None 376 | 377 | # Open file for saving full games (pgn) 378 | games_file = args.games_file.open('a') if args.games_file else sys.stdout 379 | 380 | # Run games 381 | try: 382 | runner = ArenaRunner(engines, opt, x_to_args, args.n, args.concurrency, args.games_per_encounter) 383 | arena_args = (book, limit, args.max_len, win_adj_count, win_adj_score) 384 | games_done = 0 385 | async for game_id, x, y, games, er in runner.run(arena_args): 386 | for game in games: 387 | print(game, end='\n\n', file=games_file, flush=True) 388 | if er: 389 | print('Game erred:', er, type(er)) 390 | continue 391 | 392 | results = ', '.join(g.headers['Result'] for g in games) 393 | print(f'Finished game {game_id} {x} => {y} ({results})') 394 | 395 | if data_logger: 396 | data_logger.store(x, y) 397 | 398 | games_done += 1 399 | if args.result_interval > 0 and games_done % args.result_interval == 0: 400 | await summarize(opt, steps=1) 401 | except asyncio.CancelledError: 402 | pass 403 | 404 | # Summarize final results 405 | if (await opt.size()): 406 | await summarize(opt, steps=6, restarts=10) 407 | if len(dimensions) <= 2: 408 | await opt.plot() 409 | else: 410 | print('Not enought data to summarize results.') 411 | 412 | # Close down 413 | logging.debug('Quitting engines') 414 | try: 415 | # Could also use wait here, but wait for some reason fails if the list 416 | # is empty. Why can't we just wait for nothing? 417 | await asyncio.gather(*(e.quit() for es in engines for e in es 418 | if not e.returncode.done())) 419 | except chess.engine.EngineError: 420 | pass 421 | 422 | 423 | if __name__ == '__main__': 424 | asyncio.set_event_loop_policy(chess.engine.EventLoopPolicy()) 425 | try: 426 | if hasattr(asyncio, 'run'): 427 | asyncio.run(main()) 428 | else: 429 | asyncio.get_event_loop().run_until_complete(main()) 430 | except KeyboardInterrupt: 431 | logging.debug('KeyboardInterrupt at root') 432 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU GENERAL PUBLIC LICENSE 2 | Version 3, 29 June 2007 3 | 4 | Copyright (C) 2007 Free Software Foundation, Inc. 5 | Everyone is permitted to copy and distribute verbatim copies 6 | of this license document, but changing it is not allowed. 7 | 8 | Preamble 9 | 10 | The GNU General Public License is a free, copyleft license for 11 | software and other kinds of works. 12 | 13 | The licenses for most software and other practical works are designed 14 | to take away your freedom to share and change the works. By contrast, 15 | the GNU General Public License is intended to guarantee your freedom to 16 | share and change all versions of a program--to make sure it remains free 17 | software for all its users. We, the Free Software Foundation, use the 18 | GNU General Public License for most of our software; it applies also to 19 | any other work released this way by its authors. You can apply it to 20 | your programs, too. 21 | 22 | When we speak of free software, we are referring to freedom, not 23 | price. Our General Public Licenses are designed to make sure that you 24 | have the freedom to distribute copies of free software (and charge for 25 | them if you wish), that you receive source code or can get it if you 26 | want it, that you can change the software or use pieces of it in new 27 | free programs, and that you know you can do these things. 28 | 29 | To protect your rights, we need to prevent others from denying you 30 | these rights or asking you to surrender the rights. Therefore, you have 31 | certain responsibilities if you distribute copies of the software, or if 32 | you modify it: responsibilities to respect the freedom of others. 33 | 34 | For example, if you distribute copies of such a program, whether 35 | gratis or for a fee, you must pass on to the recipients the same 36 | freedoms that you received. You must make sure that they, too, receive 37 | or can get the source code. And you must show them these terms so they 38 | know their rights. 39 | 40 | Developers that use the GNU GPL protect your rights with two steps: 41 | (1) assert copyright on the software, and (2) offer you this License 42 | giving you legal permission to copy, distribute and/or modify it. 43 | 44 | For the developers' and authors' protection, the GPL clearly explains 45 | that there is no warranty for this free software. For both users' and 46 | authors' sake, the GPL requires that modified versions be marked as 47 | changed, so that their problems will not be attributed erroneously to 48 | authors of previous versions. 49 | 50 | Some devices are designed to deny users access to install or run 51 | modified versions of the software inside them, although the manufacturer 52 | can do so. This is fundamentally incompatible with the aim of 53 | protecting users' freedom to change the software. The systematic 54 | pattern of such abuse occurs in the area of products for individuals to 55 | use, which is precisely where it is most unacceptable. Therefore, we 56 | have designed this version of the GPL to prohibit the practice for those 57 | products. If such problems arise substantially in other domains, we 58 | stand ready to extend this provision to those domains in future versions 59 | of the GPL, as needed to protect the freedom of users. 60 | 61 | Finally, every program is threatened constantly by software patents. 62 | States should not allow patents to restrict development and use of 63 | software on general-purpose computers, but in those that do, we wish to 64 | avoid the special danger that patents applied to a free program could 65 | make it effectively proprietary. To prevent this, the GPL assures that 66 | patents cannot be used to render the program non-free. 67 | 68 | The precise terms and conditions for copying, distribution and 69 | modification follow. 70 | 71 | TERMS AND CONDITIONS 72 | 73 | 0. Definitions. 74 | 75 | "This License" refers to version 3 of the GNU General Public License. 76 | 77 | "Copyright" also means copyright-like laws that apply to other kinds of 78 | works, such as semiconductor masks. 79 | 80 | "The Program" refers to any copyrightable work licensed under this 81 | License. Each licensee is addressed as "you". "Licensees" and 82 | "recipients" may be individuals or organizations. 83 | 84 | To "modify" a work means to copy from or adapt all or part of the work 85 | in a fashion requiring copyright permission, other than the making of an 86 | exact copy. The resulting work is called a "modified version" of the 87 | earlier work or a work "based on" the earlier work. 88 | 89 | A "covered work" means either the unmodified Program or a work based 90 | on the Program. 91 | 92 | To "propagate" a work means to do anything with it that, without 93 | permission, would make you directly or secondarily liable for 94 | infringement under applicable copyright law, except executing it on a 95 | computer or modifying a private copy. Propagation includes copying, 96 | distribution (with or without modification), making available to the 97 | public, and in some countries other activities as well. 98 | 99 | To "convey" a work means any kind of propagation that enables other 100 | parties to make or receive copies. Mere interaction with a user through 101 | a computer network, with no transfer of a copy, is not conveying. 102 | 103 | An interactive user interface displays "Appropriate Legal Notices" 104 | to the extent that it includes a convenient and prominently visible 105 | feature that (1) displays an appropriate copyright notice, and (2) 106 | tells the user that there is no warranty for the work (except to the 107 | extent that warranties are provided), that licensees may convey the 108 | work under this License, and how to view a copy of this License. If 109 | the interface presents a list of user commands or options, such as a 110 | menu, a prominent item in the list meets this criterion. 111 | 112 | 1. Source Code. 113 | 114 | The "source code" for a work means the preferred form of the work 115 | for making modifications to it. "Object code" means any non-source 116 | form of a work. 117 | 118 | A "Standard Interface" means an interface that either is an official 119 | standard defined by a recognized standards body, or, in the case of 120 | interfaces specified for a particular programming language, one that 121 | is widely used among developers working in that language. 122 | 123 | The "System Libraries" of an executable work include anything, other 124 | than the work as a whole, that (a) is included in the normal form of 125 | packaging a Major Component, but which is not part of that Major 126 | Component, and (b) serves only to enable use of the work with that 127 | Major Component, or to implement a Standard Interface for which an 128 | implementation is available to the public in source code form. A 129 | "Major Component", in this context, means a major essential component 130 | (kernel, window system, and so on) of the specific operating system 131 | (if any) on which the executable work runs, or a compiler used to 132 | produce the work, or an object code interpreter used to run it. 133 | 134 | The "Corresponding Source" for a work in object code form means all 135 | the source code needed to generate, install, and (for an executable 136 | work) run the object code and to modify the work, including scripts to 137 | control those activities. However, it does not include the work's 138 | System Libraries, or general-purpose tools or generally available free 139 | programs which are used unmodified in performing those activities but 140 | which are not part of the work. For example, Corresponding Source 141 | includes interface definition files associated with source files for 142 | the work, and the source code for shared libraries and dynamically 143 | linked subprograms that the work is specifically designed to require, 144 | such as by intimate data communication or control flow between those 145 | subprograms and other parts of the work. 146 | 147 | The Corresponding Source need not include anything that users 148 | can regenerate automatically from other parts of the Corresponding 149 | Source. 150 | 151 | The Corresponding Source for a work in source code form is that 152 | same work. 153 | 154 | 2. Basic Permissions. 155 | 156 | All rights granted under this License are granted for the term of 157 | copyright on the Program, and are irrevocable provided the stated 158 | conditions are met. This License explicitly affirms your unlimited 159 | permission to run the unmodified Program. The output from running a 160 | covered work is covered by this License only if the output, given its 161 | content, constitutes a covered work. This License acknowledges your 162 | rights of fair use or other equivalent, as provided by copyright law. 163 | 164 | You may make, run and propagate covered works that you do not 165 | convey, without conditions so long as your license otherwise remains 166 | in force. You may convey covered works to others for the sole purpose 167 | of having them make modifications exclusively for you, or provide you 168 | with facilities for running those works, provided that you comply with 169 | the terms of this License in conveying all material for which you do 170 | not control copyright. Those thus making or running the covered works 171 | for you must do so exclusively on your behalf, under your direction 172 | and control, on terms that prohibit them from making any copies of 173 | your copyrighted material outside their relationship with you. 174 | 175 | Conveying under any other circumstances is permitted solely under 176 | the conditions stated below. Sublicensing is not allowed; section 10 177 | makes it unnecessary. 178 | 179 | 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 180 | 181 | No covered work shall be deemed part of an effective technological 182 | measure under any applicable law fulfilling obligations under article 183 | 11 of the WIPO copyright treaty adopted on 20 December 1996, or 184 | similar laws prohibiting or restricting circumvention of such 185 | measures. 186 | 187 | When you convey a covered work, you waive any legal power to forbid 188 | circumvention of technological measures to the extent such circumvention 189 | is effected by exercising rights under this License with respect to 190 | the covered work, and you disclaim any intention to limit operation or 191 | modification of the work as a means of enforcing, against the work's 192 | users, your or third parties' legal rights to forbid circumvention of 193 | technological measures. 194 | 195 | 4. Conveying Verbatim Copies. 196 | 197 | You may convey verbatim copies of the Program's source code as you 198 | receive it, in any medium, provided that you conspicuously and 199 | appropriately publish on each copy an appropriate copyright notice; 200 | keep intact all notices stating that this License and any 201 | non-permissive terms added in accord with section 7 apply to the code; 202 | keep intact all notices of the absence of any warranty; and give all 203 | recipients a copy of this License along with the Program. 204 | 205 | You may charge any price or no price for each copy that you convey, 206 | and you may offer support or warranty protection for a fee. 207 | 208 | 5. Conveying Modified Source Versions. 209 | 210 | You may convey a work based on the Program, or the modifications to 211 | produce it from the Program, in the form of source code under the 212 | terms of section 4, provided that you also meet all of these conditions: 213 | 214 | a) The work must carry prominent notices stating that you modified 215 | it, and giving a relevant date. 216 | 217 | b) The work must carry prominent notices stating that it is 218 | released under this License and any conditions added under section 219 | 7. This requirement modifies the requirement in section 4 to 220 | "keep intact all notices". 221 | 222 | c) You must license the entire work, as a whole, under this 223 | License to anyone who comes into possession of a copy. This 224 | License will therefore apply, along with any applicable section 7 225 | additional terms, to the whole of the work, and all its parts, 226 | regardless of how they are packaged. This License gives no 227 | permission to license the work in any other way, but it does not 228 | invalidate such permission if you have separately received it. 229 | 230 | d) If the work has interactive user interfaces, each must display 231 | Appropriate Legal Notices; however, if the Program has interactive 232 | interfaces that do not display Appropriate Legal Notices, your 233 | work need not make them do so. 234 | 235 | A compilation of a covered work with other separate and independent 236 | works, which are not by their nature extensions of the covered work, 237 | and which are not combined with it such as to form a larger program, 238 | in or on a volume of a storage or distribution medium, is called an 239 | "aggregate" if the compilation and its resulting copyright are not 240 | used to limit the access or legal rights of the compilation's users 241 | beyond what the individual works permit. Inclusion of a covered work 242 | in an aggregate does not cause this License to apply to the other 243 | parts of the aggregate. 244 | 245 | 6. Conveying Non-Source Forms. 246 | 247 | You may convey a covered work in object code form under the terms 248 | of sections 4 and 5, provided that you also convey the 249 | machine-readable Corresponding Source under the terms of this License, 250 | in one of these ways: 251 | 252 | a) Convey the object code in, or embodied in, a physical product 253 | (including a physical distribution medium), accompanied by the 254 | Corresponding Source fixed on a durable physical medium 255 | customarily used for software interchange. 256 | 257 | b) Convey the object code in, or embodied in, a physical product 258 | (including a physical distribution medium), accompanied by a 259 | written offer, valid for at least three years and valid for as 260 | long as you offer spare parts or customer support for that product 261 | model, to give anyone who possesses the object code either (1) a 262 | copy of the Corresponding Source for all the software in the 263 | product that is covered by this License, on a durable physical 264 | medium customarily used for software interchange, for a price no 265 | more than your reasonable cost of physically performing this 266 | conveying of source, or (2) access to copy the 267 | Corresponding Source from a network server at no charge. 268 | 269 | c) Convey individual copies of the object code with a copy of the 270 | written offer to provide the Corresponding Source. This 271 | alternative is allowed only occasionally and noncommercially, and 272 | only if you received the object code with such an offer, in accord 273 | with subsection 6b. 274 | 275 | d) Convey the object code by offering access from a designated 276 | place (gratis or for a charge), and offer equivalent access to the 277 | Corresponding Source in the same way through the same place at no 278 | further charge. You need not require recipients to copy the 279 | Corresponding Source along with the object code. If the place to 280 | copy the object code is a network server, the Corresponding Source 281 | may be on a different server (operated by you or a third party) 282 | that supports equivalent copying facilities, provided you maintain 283 | clear directions next to the object code saying where to find the 284 | Corresponding Source. Regardless of what server hosts the 285 | Corresponding Source, you remain obligated to ensure that it is 286 | available for as long as needed to satisfy these requirements. 287 | 288 | e) Convey the object code using peer-to-peer transmission, provided 289 | you inform other peers where the object code and Corresponding 290 | Source of the work are being offered to the general public at no 291 | charge under subsection 6d. 292 | 293 | A separable portion of the object code, whose source code is excluded 294 | from the Corresponding Source as a System Library, need not be 295 | included in conveying the object code work. 296 | 297 | A "User Product" is either (1) a "consumer product", which means any 298 | tangible personal property which is normally used for personal, family, 299 | or household purposes, or (2) anything designed or sold for incorporation 300 | into a dwelling. In determining whether a product is a consumer product, 301 | doubtful cases shall be resolved in favor of coverage. For a particular 302 | product received by a particular user, "normally used" refers to a 303 | typical or common use of that class of product, regardless of the status 304 | of the particular user or of the way in which the particular user 305 | actually uses, or expects or is expected to use, the product. A product 306 | is a consumer product regardless of whether the product has substantial 307 | commercial, industrial or non-consumer uses, unless such uses represent 308 | the only significant mode of use of the product. 309 | 310 | "Installation Information" for a User Product means any methods, 311 | procedures, authorization keys, or other information required to install 312 | and execute modified versions of a covered work in that User Product from 313 | a modified version of its Corresponding Source. The information must 314 | suffice to ensure that the continued functioning of the modified object 315 | code is in no case prevented or interfered with solely because 316 | modification has been made. 317 | 318 | If you convey an object code work under this section in, or with, or 319 | specifically for use in, a User Product, and the conveying occurs as 320 | part of a transaction in which the right of possession and use of the 321 | User Product is transferred to the recipient in perpetuity or for a 322 | fixed term (regardless of how the transaction is characterized), the 323 | Corresponding Source conveyed under this section must be accompanied 324 | by the Installation Information. But this requirement does not apply 325 | if neither you nor any third party retains the ability to install 326 | modified object code on the User Product (for example, the work has 327 | been installed in ROM). 328 | 329 | The requirement to provide Installation Information does not include a 330 | requirement to continue to provide support service, warranty, or updates 331 | for a work that has been modified or installed by the recipient, or for 332 | the User Product in which it has been modified or installed. Access to a 333 | network may be denied when the modification itself materially and 334 | adversely affects the operation of the network or violates the rules and 335 | protocols for communication across the network. 336 | 337 | Corresponding Source conveyed, and Installation Information provided, 338 | in accord with this section must be in a format that is publicly 339 | documented (and with an implementation available to the public in 340 | source code form), and must require no special password or key for 341 | unpacking, reading or copying. 342 | 343 | 7. Additional Terms. 344 | 345 | "Additional permissions" are terms that supplement the terms of this 346 | License by making exceptions from one or more of its conditions. 347 | Additional permissions that are applicable to the entire Program shall 348 | be treated as though they were included in this License, to the extent 349 | that they are valid under applicable law. If additional permissions 350 | apply only to part of the Program, that part may be used separately 351 | under those permissions, but the entire Program remains governed by 352 | this License without regard to the additional permissions. 353 | 354 | When you convey a copy of a covered work, you may at your option 355 | remove any additional permissions from that copy, or from any part of 356 | it. (Additional permissions may be written to require their own 357 | removal in certain cases when you modify the work.) You may place 358 | additional permissions on material, added by you to a covered work, 359 | for which you have or can give appropriate copyright permission. 360 | 361 | Notwithstanding any other provision of this License, for material you 362 | add to a covered work, you may (if authorized by the copyright holders of 363 | that material) supplement the terms of this License with terms: 364 | 365 | a) Disclaiming warranty or limiting liability differently from the 366 | terms of sections 15 and 16 of this License; or 367 | 368 | b) Requiring preservation of specified reasonable legal notices or 369 | author attributions in that material or in the Appropriate Legal 370 | Notices displayed by works containing it; or 371 | 372 | c) Prohibiting misrepresentation of the origin of that material, or 373 | requiring that modified versions of such material be marked in 374 | reasonable ways as different from the original version; or 375 | 376 | d) Limiting the use for publicity purposes of names of licensors or 377 | authors of the material; or 378 | 379 | e) Declining to grant rights under trademark law for use of some 380 | trade names, trademarks, or service marks; or 381 | 382 | f) Requiring indemnification of licensors and authors of that 383 | material by anyone who conveys the material (or modified versions of 384 | it) with contractual assumptions of liability to the recipient, for 385 | any liability that these contractual assumptions directly impose on 386 | those licensors and authors. 387 | 388 | All other non-permissive additional terms are considered "further 389 | restrictions" within the meaning of section 10. If the Program as you 390 | received it, or any part of it, contains a notice stating that it is 391 | governed by this License along with a term that is a further 392 | restriction, you may remove that term. If a license document contains 393 | a further restriction but permits relicensing or conveying under this 394 | License, you may add to a covered work material governed by the terms 395 | of that license document, provided that the further restriction does 396 | not survive such relicensing or conveying. 397 | 398 | If you add terms to a covered work in accord with this section, you 399 | must place, in the relevant source files, a statement of the 400 | additional terms that apply to those files, or a notice indicating 401 | where to find the applicable terms. 402 | 403 | Additional terms, permissive or non-permissive, may be stated in the 404 | form of a separately written license, or stated as exceptions; 405 | the above requirements apply either way. 406 | 407 | 8. Termination. 408 | 409 | You may not propagate or modify a covered work except as expressly 410 | provided under this License. Any attempt otherwise to propagate or 411 | modify it is void, and will automatically terminate your rights under 412 | this License (including any patent licenses granted under the third 413 | paragraph of section 11). 414 | 415 | However, if you cease all violation of this License, then your 416 | license from a particular copyright holder is reinstated (a) 417 | provisionally, unless and until the copyright holder explicitly and 418 | finally terminates your license, and (b) permanently, if the copyright 419 | holder fails to notify you of the violation by some reasonable means 420 | prior to 60 days after the cessation. 421 | 422 | Moreover, your license from a particular copyright holder is 423 | reinstated permanently if the copyright holder notifies you of the 424 | violation by some reasonable means, this is the first time you have 425 | received notice of violation of this License (for any work) from that 426 | copyright holder, and you cure the violation prior to 30 days after 427 | your receipt of the notice. 428 | 429 | Termination of your rights under this section does not terminate the 430 | licenses of parties who have received copies or rights from you under 431 | this License. If your rights have been terminated and not permanently 432 | reinstated, you do not qualify to receive new licenses for the same 433 | material under section 10. 434 | 435 | 9. Acceptance Not Required for Having Copies. 436 | 437 | You are not required to accept this License in order to receive or 438 | run a copy of the Program. Ancillary propagation of a covered work 439 | occurring solely as a consequence of using peer-to-peer transmission 440 | to receive a copy likewise does not require acceptance. However, 441 | nothing other than this License grants you permission to propagate or 442 | modify any covered work. These actions infringe copyright if you do 443 | not accept this License. Therefore, by modifying or propagating a 444 | covered work, you indicate your acceptance of this License to do so. 445 | 446 | 10. Automatic Licensing of Downstream Recipients. 447 | 448 | Each time you convey a covered work, the recipient automatically 449 | receives a license from the original licensors, to run, modify and 450 | propagate that work, subject to this License. You are not responsible 451 | for enforcing compliance by third parties with this License. 452 | 453 | An "entity transaction" is a transaction transferring control of an 454 | organization, or substantially all assets of one, or subdividing an 455 | organization, or merging organizations. If propagation of a covered 456 | work results from an entity transaction, each party to that 457 | transaction who receives a copy of the work also receives whatever 458 | licenses to the work the party's predecessor in interest had or could 459 | give under the previous paragraph, plus a right to possession of the 460 | Corresponding Source of the work from the predecessor in interest, if 461 | the predecessor has it or can get it with reasonable efforts. 462 | 463 | You may not impose any further restrictions on the exercise of the 464 | rights granted or affirmed under this License. For example, you may 465 | not impose a license fee, royalty, or other charge for exercise of 466 | rights granted under this License, and you may not initiate litigation 467 | (including a cross-claim or counterclaim in a lawsuit) alleging that 468 | any patent claim is infringed by making, using, selling, offering for 469 | sale, or importing the Program or any portion of it. 470 | 471 | 11. Patents. 472 | 473 | A "contributor" is a copyright holder who authorizes use under this 474 | License of the Program or a work on which the Program is based. The 475 | work thus licensed is called the contributor's "contributor version". 476 | 477 | A contributor's "essential patent claims" are all patent claims 478 | owned or controlled by the contributor, whether already acquired or 479 | hereafter acquired, that would be infringed by some manner, permitted 480 | by this License, of making, using, or selling its contributor version, 481 | but do not include claims that would be infringed only as a 482 | consequence of further modification of the contributor version. For 483 | purposes of this definition, "control" includes the right to grant 484 | patent sublicenses in a manner consistent with the requirements of 485 | this License. 486 | 487 | Each contributor grants you a non-exclusive, worldwide, royalty-free 488 | patent license under the contributor's essential patent claims, to 489 | make, use, sell, offer for sale, import and otherwise run, modify and 490 | propagate the contents of its contributor version. 491 | 492 | In the following three paragraphs, a "patent license" is any express 493 | agreement or commitment, however denominated, not to enforce a patent 494 | (such as an express permission to practice a patent or covenant not to 495 | sue for patent infringement). To "grant" such a patent license to a 496 | party means to make such an agreement or commitment not to enforce a 497 | patent against the party. 498 | 499 | If you convey a covered work, knowingly relying on a patent license, 500 | and the Corresponding Source of the work is not available for anyone 501 | to copy, free of charge and under the terms of this License, through a 502 | publicly available network server or other readily accessible means, 503 | then you must either (1) cause the Corresponding Source to be so 504 | available, or (2) arrange to deprive yourself of the benefit of the 505 | patent license for this particular work, or (3) arrange, in a manner 506 | consistent with the requirements of this License, to extend the patent 507 | license to downstream recipients. "Knowingly relying" means you have 508 | actual knowledge that, but for the patent license, your conveying the 509 | covered work in a country, or your recipient's use of the covered work 510 | in a country, would infringe one or more identifiable patents in that 511 | country that you have reason to believe are valid. 512 | 513 | If, pursuant to or in connection with a single transaction or 514 | arrangement, you convey, or propagate by procuring conveyance of, a 515 | covered work, and grant a patent license to some of the parties 516 | receiving the covered work authorizing them to use, propagate, modify 517 | or convey a specific copy of the covered work, then the patent license 518 | you grant is automatically extended to all recipients of the covered 519 | work and works based on it. 520 | 521 | A patent license is "discriminatory" if it does not include within 522 | the scope of its coverage, prohibits the exercise of, or is 523 | conditioned on the non-exercise of one or more of the rights that are 524 | specifically granted under this License. You may not convey a covered 525 | work if you are a party to an arrangement with a third party that is 526 | in the business of distributing software, under which you make payment 527 | to the third party based on the extent of your activity of conveying 528 | the work, and under which the third party grants, to any of the 529 | parties who would receive the covered work from you, a discriminatory 530 | patent license (a) in connection with copies of the covered work 531 | conveyed by you (or copies made from those copies), or (b) primarily 532 | for and in connection with specific products or compilations that 533 | contain the covered work, unless you entered into that arrangement, 534 | or that patent license was granted, prior to 28 March 2007. 535 | 536 | Nothing in this License shall be construed as excluding or limiting 537 | any implied license or other defenses to infringement that may 538 | otherwise be available to you under applicable patent law. 539 | 540 | 12. No Surrender of Others' Freedom. 541 | 542 | If conditions are imposed on you (whether by court order, agreement or 543 | otherwise) that contradict the conditions of this License, they do not 544 | excuse you from the conditions of this License. If you cannot convey a 545 | covered work so as to satisfy simultaneously your obligations under this 546 | License and any other pertinent obligations, then as a consequence you may 547 | not convey it at all. For example, if you agree to terms that obligate you 548 | to collect a royalty for further conveying from those to whom you convey 549 | the Program, the only way you could satisfy both those terms and this 550 | License would be to refrain entirely from conveying the Program. 551 | 552 | 13. Use with the GNU Affero General Public License. 553 | 554 | Notwithstanding any other provision of this License, you have 555 | permission to link or combine any covered work with a work licensed 556 | under version 3 of the GNU Affero General Public License into a single 557 | combined work, and to convey the resulting work. The terms of this 558 | License will continue to apply to the part which is the covered work, 559 | but the special requirements of the GNU Affero General Public License, 560 | section 13, concerning interaction through a network will apply to the 561 | combination as such. 562 | 563 | 14. Revised Versions of this License. 564 | 565 | The Free Software Foundation may publish revised and/or new versions of 566 | the GNU General Public License from time to time. Such new versions will 567 | be similar in spirit to the present version, but may differ in detail to 568 | address new problems or concerns. 569 | 570 | Each version is given a distinguishing version number. If the 571 | Program specifies that a certain numbered version of the GNU General 572 | Public License "or any later version" applies to it, you have the 573 | option of following the terms and conditions either of that numbered 574 | version or of any later version published by the Free Software 575 | Foundation. If the Program does not specify a version number of the 576 | GNU General Public License, you may choose any version ever published 577 | by the Free Software Foundation. 578 | 579 | If the Program specifies that a proxy can decide which future 580 | versions of the GNU General Public License can be used, that proxy's 581 | public statement of acceptance of a version permanently authorizes you 582 | to choose that version for the Program. 583 | 584 | Later license versions may give you additional or different 585 | permissions. However, no additional obligations are imposed on any 586 | author or copyright holder as a result of your choosing to follow a 587 | later version. 588 | 589 | 15. Disclaimer of Warranty. 590 | 591 | THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 592 | APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT 593 | HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY 594 | OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, 595 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 596 | PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM 597 | IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF 598 | ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 599 | 600 | 16. Limitation of Liability. 601 | 602 | IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 603 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS 604 | THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY 605 | GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE 606 | USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF 607 | DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 608 | PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), 609 | EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF 610 | SUCH DAMAGES. 611 | 612 | 17. Interpretation of Sections 15 and 16. 613 | 614 | If the disclaimer of warranty and limitation of liability provided 615 | above cannot be given local legal effect according to their terms, 616 | reviewing courts shall apply local law that most closely approximates 617 | an absolute waiver of all civil liability in connection with the 618 | Program, unless a warranty or assumption of liability accompanies a 619 | copy of the Program in return for a fee. 620 | 621 | END OF TERMS AND CONDITIONS 622 | 623 | How to Apply These Terms to Your New Programs 624 | 625 | If you develop a new program, and you want it to be of the greatest 626 | possible use to the public, the best way to achieve this is to make it 627 | free software which everyone can redistribute and change under these terms. 628 | 629 | To do so, attach the following notices to the program. It is safest 630 | to attach them to the start of each source file to most effectively 631 | state the exclusion of warranty; and each file should have at least 632 | the "copyright" line and a pointer to where the full notice is found. 633 | 634 | 635 | Copyright (C) 636 | 637 | This program is free software: you can redistribute it and/or modify 638 | it under the terms of the GNU General Public License as published by 639 | the Free Software Foundation, either version 3 of the License, or 640 | (at your option) any later version. 641 | 642 | This program is distributed in the hope that it will be useful, 643 | but WITHOUT ANY WARRANTY; without even the implied warranty of 644 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 645 | GNU General Public License for more details. 646 | 647 | You should have received a copy of the GNU General Public License 648 | along with this program. If not, see . 649 | 650 | Also add information on how to contact you by electronic and paper mail. 651 | 652 | If the program does terminal interaction, make it output a short 653 | notice like this when it starts in an interactive mode: 654 | 655 | Copyright (C) 656 | This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. 657 | This is free software, and you are welcome to redistribute it 658 | under certain conditions; type `show c' for details. 659 | 660 | The hypothetical commands `show w' and `show c' should show the appropriate 661 | parts of the General Public License. Of course, your program's commands 662 | might be different; for a GUI interface, you would use an "about box". 663 | 664 | You should also get your employer (if you work as a programmer) or school, 665 | if any, to sign a "copyright disclaimer" for the program, if necessary. 666 | For more information on this, and how to apply and follow the GNU GPL, see 667 | . 668 | 669 | The GNU General Public License does not permit incorporating your program 670 | into proprietary programs. If your program is a subroutine library, you 671 | may consider it more useful to permit linking proprietary applications with 672 | the library. If this is what you want to do, use the GNU Lesser General 673 | Public License instead of this License. But first, please read 674 | . 675 | --------------------------------------------------------------------------------