mdp.py 26.7 KB
Newer Older
Steven Cordwell's avatar
Steven Cordwell committed
1
2
3
# -*- coding: utf-8 -*-
"""
Copyright (c) 2011, 2012 Steven Cordwell
Steven Cordwell's avatar
Steven Cordwell committed
4
5
6
7
Copyright (c) 2009, Iadine Chadès
Copyright (c) 2009, Marie-Josée Cros
Copyright (c) 2009, Frédérick Garcia
Copyright (c) 2009, Régis Sabbadin
Steven Cordwell's avatar
Steven Cordwell committed
8
9
10

All rights reserved.

Steven Cordwell's avatar
Steven Cordwell committed
11
12
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
Steven Cordwell's avatar
Steven Cordwell committed
13
14
15
16
17
18

  * Redistributions of source code must retain the above copyright notice, this
    list of conditions and the following disclaimer.
  * Redistributions in binary form must reproduce the above copyright notice,
    this list of conditions and the following disclaimer in the documentation
    and/or other materials provided with the distribution.
Steven Cordwell's avatar
Steven Cordwell committed
19
20
21
  * Neither the name of the <ORGANIZATION> nor the names of its contributors
    may be used to endorse or promote products derived from this software
    without specific prior written permission.
Steven Cordwell's avatar
Steven Cordwell committed
22
23
24
25

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
Steven Cordwell's avatar
Steven Cordwell committed
26
27
28
29
30
31
32
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
Steven Cordwell's avatar
Steven Cordwell committed
33
34
"""

Steven Cordwell's avatar
Steven Cordwell committed
35
36
from numpy import abs, array, diag, matrix, ndarray, ones, zeros
from numpy.random import rand
Steven Cordwell's avatar
Steven Cordwell committed
37
from math import ceil, log, sqrt
Steven Cordwell's avatar
Steven Cordwell committed
38
39
from random import randint, random
from scipy.sparse import csr_matrix as sparse
Steven Cordwell's avatar
Steven Cordwell committed
40
41
from time import time

Steven Cordwell's avatar
Steven Cordwell committed
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
mdperr = {
"mat_nonneg" :
    "PyMDPtoolbox: Probabilities must be non-negative.",
"mat_square" :
    "PyMDPtoolbox: The matrix must be square.",
"mat_stoch" :
    "PyMDPtoolbox: Rows of the matrix must sum to one (1).",
"mask_numpy" :
    "PyMDPtoolbox: mask must be a numpy array or matrix; i.e. type(mask) is "
    "ndarray or type(mask) is matrix.", 
"mask_SbyS" : 
    "PyMDPtoolbox: The mask must have shape SxS; i.e. mask.shape = (S, S).",
"obj_shape" :
    "PyMDPtoolbox: Object arrays for transition probabilities and rewards "
    "must have only 1 dimension: the number of actions A. Each element of "
    "the object array contains an SxS ndarray or matrix.",
"obj_square" :
    "PyMDPtoolbox: Each element of an object array for transition "
    "probabilities and rewards must contain an SxS ndarray or matrix; i.e. "
    "P[a].shape = (S, S) or R[a].shape = (S, S).",
"P_ndarray" :
    "PyMDPtoolbox: The transition probabilities must be in a numpy array; "
    "i.e. type(P) is ndarray.",
"P_shape" :
    "PyMDPtoolbox: The transition probability array must have the shape "
    "(A, S, S)  with S : number of states greater than 0 and A : number of "
    "actions greater than 0. i.e. R.shape = (A, S, S)",
"PR_incompat" :
    "PyMDPtoolbox: Incompatibility between P and R dimensions.",
"prob_in01" :
    "PyMDPtoolbox: Probability p must be in [0; 1].",
"R_ndarray" :
    "PyMDPtoolbox: The rewards must be in a numpy array; i.e. type(R) is "
    "ndarray.",
"R_shape" :
    "PyMDPtoolbox: The reward matrix R must be an array of shape (A, S, S) or "
    "(S, A) with S : number of states greater than 0 and A : number of actions "
    "greater than 0. i.e. R.shape = (S, A) or (A, S, S).",
"R_gt_0" :
    "PyMDPtoolbox: The rewards must be greater than 0.",
"S_gt_1" :
    "PyMDPtoolbox: Number of states S must be greater than 1.",
"SA_gt_1" : 
    "PyMDPtoolbox: The number of states S and the number of actions A must be "
    "greater than 1."
}

def exampleForest(S=3, r1=4, r2=2, p=0.1):
    """
    Generates a Markov Decision Process example based on a simple forest
    management.
    
    See the related documentation for more detail.
    
    Parameters
    ---------
    S : number of states (> 0), optional (default 3)
    r1 : reward when forest is in the oldest state and action Wait is performed,
        optional (default 4)
    r2 : reward when forest is in the oldest state and action Cut is performed, 
        optional (default 2)
    p : probability of wild fire occurence, in ]0, 1[, optional (default 0.1)
    
    Evaluation
    ----------
    P : transition probability matrix (A, S, S)
    R : reward matrix (S, A)
    """
    if (S <= 1):
        raise ValueError(mdperr["S_gt_1"])
    if (r1 <= 0) or (r2 <= 0):
        raise ValueError(mdperr["R_gt_0"])
    if (p < 0 or p > 1):
        raise ValueError(mdperr["prob_in01"])
    
    # Definition of Transition matrix P(:,:,1) associated to action Wait (action 1) and
    # P(:,:,2) associated to action Cut (action 2)
    #             | p 1-p 0.......0  |                  | 1 0..........0 |
    #             | .  0 1-p 0....0  |                  | . .          . |
    #  P(:,:,1) = | .  .  0  .       |  and P(:,:,2) =  | . .          . |
    #             | .  .        .    |                  | . .          . |
    #             | .  .         1-p |                  | . .          . |
    #             | p  0  0....0 1-p |                  | 1 0..........0 |
    P = zeros((2, S, S))
    P[0, :, :] = (1 - p) * diag(ones(S - 1), 1)
    P[0, :, 0] = p
    P[0, S - 1, S - 1] = (1 - p)
    P[1, :, :] = zeros((S, S))
    P[1, :, 0] = 1
    
    # Definition of Reward matrix R1 associated to action Wait and 
    # R2 associated to action Cut
    #           | 0  |                   | 0  |
    #           | .  |                   | 1  |
    #  R(:,1) = | .  |  and     R(:,2) = | .  |	
    #           | .  |                   | .  |
    #           | 0  |                   | 1  |                   
    #           | r1 |                   | r2 |
    R = zeros((S, 2))
    R[S - 1, 0] = r1
    R[:, 1] = ones(S)
    R[0, 1] = 0
    R[S - 1, 1] = r2
    
    return (P, R)

def exampleRand(S, A, is_sparse=False, mask=None):
    """Generates a random Markov Decision Process.
    
    Parameters
    ----------
    S : number of states (> 0)
    A : number of actions (> 0)
    is_sparse : false to have matrices in plain format, true to have sparse
        matrices optional (default false).
    mask : matrix with 0 and 1 (0 indicates a place for a zero
           probability), optional (SxS) (default, random )
    
    Returns
    ----------
    P : transition probability matrix (SxSxA)
    R : reward matrix (SxSxA)
    """
    if (S < 1 or A < 1):
        raise ValueError(mdperr["SA_gt_1"])
    
    try:
        if (mask != None) and ((mask.shape[0] != S) or (mask.shape[1] != S)):
            raise ValueError(mdperr["mask_SbyS"])
    except AttributeError:
        raise TypeError(mdperr["mask_numpy"])
    
    if mask == None:
        mask = rand(S, S)
        r = random()
        mask[mask < r] = 0
        mask[mask >= r] = 1
        for s in range(S):
            if mask[s, :].sum() == 0:
                mask[s, randint(0, S - 1)] = 1
    
    if is_sparse:
        # definition of transition matrix : square stochastic matrix
        P = zeros((A,), dtype=object)
        # definition of reward matrix (values between -1 and +1)
        R = zeros((A,), dtype=object)
        for a in range(A):
            PP = mask * rand(S, S)
            for s in range(S):
                PP[s, :] = PP[s, :] / PP[s, :].sum()
            P[a] = sparse(PP)
            R[a] = sparse(mask * (2 * rand(S, S) - ones((S, S))))
    else:
        # definition of transition matrix : square stochastic matrix
        P = zeros((A, S, S))
        # definition of reward matrix (values between -1 and +1)
        R = zeros((A, S, S))
        for a in range(A):
            P[a, :, :] = mask * rand(S, S)
            for s in range(S):
                P[a, s, :] = P[a, s, :] / P[a, s, :].sum()
            R[a, :, :] = mask * (2 * rand(S, S) - ones((S, S)))
    
    return (P, R)

Steven Cordwell's avatar
Steven Cordwell committed
207
class MDP():
Steven Cordwell's avatar
Steven Cordwell committed
208
    """The Markov Decision Problem Toolbox."""
Steven Cordwell's avatar
Steven Cordwell committed
209
210
211
212
213
    
    def __init__(self):
        """"""
        self.verbose = False
    
Steven Cordwell's avatar
Steven Cordwell committed
214
    def bellmanOperator(self):
Steven Cordwell's avatar
Steven Cordwell committed
215
216
        """
        Applies the Bellman operator on the value function.
Steven Cordwell's avatar
Steven Cordwell committed
217
        
Steven Cordwell's avatar
Steven Cordwell committed
218
219
220
221
222
223
        Updates the value function and the Vprev-improving policy.
        
        Updates
        -------
        self.value : new value function
        self.policy : Vprev-improving policy
Steven Cordwell's avatar
Steven Cordwell committed
224
225
226
        """
        Q = matrix(zeros((self.S, self.A)))
        for aa in range(self.A):
Steven Cordwell's avatar
Steven Cordwell committed
227
            Q[:, aa] = self.R[:, aa] + (self.discount * self.P[aa] * self.value)
Steven Cordwell's avatar
Steven Cordwell committed
228
229
230
231
232
233
        
        # update the value and policy
        self.value = Q.max(axis=1)
        self.policy = Q.argmax(axis=1)
    
    def check(self, P, R):
Steven Cordwell's avatar
Steven Cordwell committed
234
        """Checks if the matrices P and R define a Markov Decision Process.
Steven Cordwell's avatar
Steven Cordwell committed
235
236
        
        Let S = number of states, A = number of actions
Steven Cordwell's avatar
Steven Cordwell committed
237
238
239
240
241
242
243
244
245
246
247
248
249
250
            The transition matrix P must be on the shape (A, S, S) and P[a,:,:]
            must be stochastic.
            The reward matrix R must be on the shape (A, S, S) or (S, A).
            Raises an error if P and R do not define a MDP.
        
        Parameters
        ---------
        P : transition matrix (A, S, S)
            P could be an array with 3 dimensions or a object array (A, ),
            each cell containing a matrix (S, S) possibly sparse
        R : reward matrix (A, S, S) or (S, A)
            R could be an array with 3 dimensions (SxSxA) or a object array
            (A, ), each cell containing a sparse matrix (S, S) or a 2D
            array(S, A) possibly sparse  
Steven Cordwell's avatar
Steven Cordwell committed
251
252
253
254
255
256
257
258
259
        """
        
        # Check of P
        # tranitions must be a numpy array either an AxSxS ndarray (with any 
        # dtype other than "object"); or, a 1xA ndarray with a "object" dtype, 
        # and each element containing an SxS array. An AxSxS array will be
        # be converted to an object array. A numpy object array is similar to a
        # MATLAB cell array.
        if (not type(P) is ndarray):
Steven Cordwell's avatar
Steven Cordwell committed
260
261
            raise TypeError(mdperr["P_ndarray"])
        
Steven Cordwell's avatar
Steven Cordwell committed
262
        if (not type(R) is ndarray):
Steven Cordwell's avatar
Steven Cordwell committed
263
264
265
266
267
268
269
            raise TypeError(mdperr["R_ndarray"])
            
        if (P.dtype == object):
            if (P.ndim > 1):
                raise ValueError(mdperr["obj_shape"])
            else:
                P_is_object = True
Steven Cordwell's avatar
Steven Cordwell committed
270
        else:
Steven Cordwell's avatar
Steven Cordwell committed
271
272
273
274
            if (P.ndim != 3):
                raise ValueError(mdperr["P_shape"])
            else:
                P_is_object = False
Steven Cordwell's avatar
Steven Cordwell committed
275
            
Steven Cordwell's avatar
Steven Cordwell committed
276
277
278
279
280
        if (R.dtype == object):
            if (R.ndim > 1):
                raise ValueError(mdperr["obj_shape"])
            else:
                R_is_object = True
Steven Cordwell's avatar
Steven Cordwell committed
281
        else:
Steven Cordwell's avatar
Steven Cordwell committed
282
283
284
285
            if (not R.ndim in (2, 3)):
                raise ValueError(mdperr["R_shape"])
            else:
                R_is_object = False
Steven Cordwell's avatar
Steven Cordwell committed
286
287
288
289
290
        
        if P_is_object:
            aP = P.shape[0]
            sP0 = P[0].shape[0]
            sP1 = P[0].shape[1]
Steven Cordwell's avatar
Steven Cordwell committed
291
            # check to see that the other object array elements are the same shape
Steven Cordwell's avatar
Steven Cordwell committed
292
293
294
295
            for aa in range(1, aP):
                sP0aa = P[aa].shape[0]
                sP1aa = P[aa].shape[1]
                if ((sP0aa != sP0) or (sP1aa != sP1)):
Steven Cordwell's avatar
Steven Cordwell committed
296
                    raise ValueError(mdperr["obj_square"])
Steven Cordwell's avatar
Steven Cordwell committed
297
298
299
300
        else:
            aP, sP0, sP1 = P.shape
        
        if ((sP0 < 1) or (aP < 1) or (sP0 != sP1)):
Steven Cordwell's avatar
Steven Cordwell committed
301
302
303
304
305
            raise ValueError(mdperr["P_shape"])
        
        for aa in range(aP):
            if P_is_object:
                self.checkSquareStochastic(P[aa])
Steven Cordwell's avatar
Steven Cordwell committed
306
            else:
Steven Cordwell's avatar
Steven Cordwell committed
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
                self.checkSquareStochastic(P[aa, :, :])
            aa = aa + 1
        
        if R_is_object:
            aR = R.shape[0]
            sR0 = R[0].shape[0]
            sR1 = R[0].shape[1]
            # check to see that the other object array elements are the same shape
            for aa in range(1, aR):
                sR0aa = R[aa].shape[0]
                sR1aa = R[aa].shape[1]
                if ((sR0aa != sR0) or (sR1aa != sR1)):
                    raise ValueError(mdperr["obj_square"])
        elif (R.ndim == 3):
            aR, sR0, sR1 = R.shape
        else:
            sR0, aR = R.shape
            sR1 = sR0
Steven Cordwell's avatar
Steven Cordwell committed
325
        
Steven Cordwell's avatar
Steven Cordwell committed
326
327
328
329
330
        if ((sR0 < 1) or (aR < 1) or (sR0 != sR1)):
            raise ValueError(mdperr["R_shape"])
                
        if (sP0 != sR0) or (aP != aR):
            raise ValueError(mdperr["PR_incompat"])
Steven Cordwell's avatar
Steven Cordwell committed
331
332
333
334
    
    def checkSquareStochastic(self, Z):
        """Check if Z is a square stochastic matrix
        
Steven Cordwell's avatar
Steven Cordwell committed
335
336
        Arguments
        --------------------------------------------------------------
Steven Cordwell's avatar
Steven Cordwell committed
337
            Z = a numpy ndarray SxS, possibly sparse (csr_matrix)
Steven Cordwell's avatar
Steven Cordwell committed
338
339
        Evaluation
        -------------------------------------------------------------
Steven Cordwell's avatar
Steven Cordwell committed
340
341
342
343
            error_msg = error message or None if correct
        """
        s1, s2 = Z.shape
        if (s1 != s2):
Steven Cordwell's avatar
Steven Cordwell committed
344
           raise ValueError(mdperr["mat_square"])
Steven Cordwell's avatar
Steven Cordwell committed
345
        elif (abs(Z.sum(axis=1) - ones(s2))).max() > 10**(-12):
Steven Cordwell's avatar
Steven Cordwell committed
346
347
348
349
350
            raise ValueError(mdperr["mat_stoch"])
        elif ((type(Z) is ndarray) or (type(Z) is matrix)) and (Z < 0).any():
            raise ValueError(mdperr["mat_nonneg"])
        elif (type(Z) is sparse) and (Z.data < 0).any():
            raise ValueError(mdperr["mat_nonneg"]) 
Steven Cordwell's avatar
Steven Cordwell committed
351
352
        else:
            return(None)
Steven Cordwell's avatar
Steven Cordwell committed
353
    
Steven Cordwell's avatar
Steven Cordwell committed
354
355
356
357
    def computePpolicyPRpolicy(self):
        """Computes the transition matrix and the reward matrix for a policy.
        """
        pass    
Steven Cordwell's avatar
Steven Cordwell committed
358
359
360
361
    
    def computePR(self, P, R):
        """Computes the reward for the system in one state chosing an action
        
Steven Cordwell's avatar
Steven Cordwell committed
362
363
        Arguments
        --------------------------------------------------------------
Steven Cordwell's avatar
Steven Cordwell committed
364
365
366
367
368
369
370
371
        Let S = number of states, A = number of actions
            P(SxSxA)  = transition matrix 
                P could be an array with 3 dimensions or  a cell array (1xA), 
                each cell containing a matrix (SxS) possibly sparse
            R(SxSxA) or (SxA) = reward matrix
                R could be an array with 3 dimensions (SxSxA) or  a cell array 
                (1xA), each cell containing a sparse matrix (SxS) or a 2D 
                array(SxA) possibly sparse  
Steven Cordwell's avatar
Steven Cordwell committed
372
373
        Evaluation
        -------------------------------------------------------------
Steven Cordwell's avatar
Steven Cordwell committed
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
            PR(SxA)   = reward matrix
        """
        # make P be an object array with (S, S) shaped array elements
        if (P.dtype == object):
            self.P = P
            self.A = self.P.shape[0]
            self.S = self.P[0].shape[0]
        else: # convert to an object array
            self.A = P.shape[0]
            self.S = P.shape[1]
            self.P = zeros(self.A, dtype=object)
            for aa in range(self.A):
                self.P[aa] = P[aa, :, :]
        
        # make R have the shape (S, A)
        if ((R.ndim == 2) and (not R.dtype is object)):
            # R already has shape (S, A)
            self.R = R
        else: 
            # R has shape (A, S, S) or object shaped (A,) with each element
            # shaped (S, S)
            self.R = zeros((self.S, self.A))
            if (R.dtype is object):
                for aa in range(self.A):
                    self.R[:, aa] = sum(P[aa] * R[aa], 2)
            else:
                for aa in range(self.A):
                    self.R[:, aa] = sum(P[aa] * R[aa, :, :], 2)
        
        # convert the arrays to numpy matrices
        for aa in range(self.A):
            if (type(self.P[aa]) is ndarray):
                self.P[aa] = matrix(self.P[aa])
        if (type(self.R) is ndarray):
            self.R = matrix(self.R)
    
Steven Cordwell's avatar
Steven Cordwell committed
410
    def getSpan(self, W):
Steven Cordwell's avatar
Steven Cordwell committed
411
412
413
414
415
        """Returns the span of W
        
        sp(W) = max W(s) - min W(s)
        """
        return (W.max() - W.min())
Steven Cordwell's avatar
Steven Cordwell committed
416
    
Steven Cordwell's avatar
Steven Cordwell committed
417
418
419
420
421
422
423
    def setSilent(self):
        """Ask for running resolution functions of the MDP Toolbox in silent
        mode.
        """
        self.verbose = False
    
    def setVerbose(self):
Steven Cordwell's avatar
Steven Cordwell committed
424
425
426
        """Ask for running resolution functions of the MDP Toolbox in verbose
        mode.
        """
Steven Cordwell's avatar
Steven Cordwell committed
427
        self.verbose = True
Steven Cordwell's avatar
Steven Cordwell committed
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481

class FiniteHorizon(MDP):
    """Resolution of finite-horizon MDP with backwards induction.
    """
    pass

class LP(MDP):
    """Resolution of discounted MDP with linear programming.
    """
    pass

class PolicyIteration(MDP):
    """Resolution of discounted MDP with policy iteration algorithm.
    """
    pass

class PolicyIterationModified(MDP):
    """Resolution of discounted MDP with modified policy iteration algorithm.
    """
    pass

class QLearning(MDP):
    """Evaluation of the matrix Q, using the Q learning algorithm.
    """
    
    def __init__(self, transitions, reward, discount, n_iter=10000):
        """Evaluation of the matrix Q, using the Q learning algorithm
        
        Arguments
        -----------------------------------------------------------------------
        Let S = number of states, A = number of actions
        transitions(SxSxA) = transition matrix 
            P could be an array with 3 dimensions or a cell array (1xA), each
            cell containing a sparse matrix (SxS)
        
        reward(SxSxA) or (SxA) = reward matrix
            R could be an array with 3 dimensions (SxSxA) or a cell array
            (1xA), each cell containing a sparse matrix (SxS) or a 2D
            array(SxA) possibly sparse
        
        discount = discount rate in ]0; 1[
        
        n_iter(optional) = number of iterations to execute.
            Default value = 10000; it is an integer greater than the default
            value.
        Evaluation
        -----------------------------------------------------------------------
        Q(SxA) = learned Q matrix 
        value(S)   = learned value function.
        policy(S) = learned optimal policy.
        mean_discrepancy(N/100) = vector of V discrepancy mean over 100 iterations
            Then the length of this vector for the default value of N is 100.
        """
        
Steven Cordwell's avatar
Steven Cordwell committed
482
483
        MDP.__init__()
        
Steven Cordwell's avatar
Steven Cordwell committed
484
485
486
487
488
489
        # Check of arguments
        if (discount <= 0) or (discount >= 1):
            raise ValueError("MDP Toolbox Error: Discount rate must be in ]0,1[")
        elif (n_iter < 10000):
            raise ValueError("MDP Toolbox Error: n_iter must be greater than 10000")
        
Steven Cordwell's avatar
Steven Cordwell committed
490
        self.check(transitions, reward)
Steven Cordwell's avatar
Steven Cordwell committed
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
        
        self.computePR(transitions, reward)
        
        self.discount = discount
        
        self.n_iter = n_iter
        
        # Initialisations
        self.Q = zeros((self.S, self.A))
        #self.dQ = zeros(self.S, self.A)
        self.mean_discrepancy = []
        self.discrepancy = zeros((self.S, 100))
        
        self.time = None
        
    def iterate(self):
        """
        """
        self.time = time()
        
        # initial state choice
        s = randint(0, self.S - 1)
        
        for n in range(self.n_iter):
            
            # Reinitialisation of trajectories every 100 transitions
            if ((n % 100) == 0):
                s = randint(1, self.S)
            
            # Action choice : greedy with increasing probability
            # probability 1-(1/log(n+2)) can be changed
            pn = random()
            if (pn < (1 - (1 / log(n + 2)))):
                optimal_action, a = self.Q[s, ].max()
            else:
                a = randint(0, self.A - 1)
              
            # Simulating next state s_new and reward associated to <s,s_new,a>
            p_s_new = random()
            p = 0
            s_new = 0
            while ((p < p_s_new) and (s_new < s)):
                s_new = s_new + 1
                if (type(self.P) is object):
                    p = p + self.P[a][s, s_new]
                else:
                    p = p + self.P[a, s, s_new]
                
            if (type(self.R) is object):
                r = self.R[a][s, s_new]
            elif (self.R.ndim == 3):
                r = self.R(a, s, s_new)
            else:
                r = self.R(s, a)
                
            # Updating the value of Q   
            # Decaying update coefficient (1/sqrt(n+2)) can be changed
            delta = r + self.discount * self.Q[s_new, ].max() - self.Q[s, a]
            dQ = (1 / sqrt(n + 2)) * delta
            self.Q[s, a] = self.Q[s, a] + dQ
            
            # current state is updated
            s = s_new
            
            # Computing and saving maximal values of the Q variation
            self.discrepancy[(n % 100) + 1, ] = abs(dQ)
            
            # Computing means all over maximal Q variations values
            if ((n % 100) == 99):
                self.mean_discrepancy.append(self.discrepancy.mean(1))
                self.discrepancy = zeros((self.S, 100))
            
            # compute the value function and the policy
            self.value = self.Q.max(axis=1)
            self.policy = self.Q.argmax(axis=1)
            
            self.time = time() - self.time

class RelativeValueIteration(MDP):
    """Resolution of MDP with average reward with relative value iteration
Steven Cordwell's avatar
Steven Cordwell committed
571
572
    algorithm.
    """
Steven Cordwell's avatar
Steven Cordwell committed
573
574
575
576
577
578
    pass

class ValueIteration(MDP):
    """
    Resolve a discounted Markov Decision Problem with value iteration.
    """
Steven Cordwell's avatar
Steven Cordwell committed
579
    
Steven Cordwell's avatar
Steven Cordwell committed
580
    def __init__(self, transitions, reward, discount, epsilon=0.01, max_iter=1000, initial_value=0):
Steven Cordwell's avatar
Steven Cordwell committed
581
582
        """Resolution of discounted MDP with value iteration algorithm.
        
Steven Cordwell's avatar
Steven Cordwell committed
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
        Arguments
        ---------
        Let S = number of states, A = number of actions.
        
        transitions = transition matrix 
            P could be a numpy ndarray with 3 dimensions (AxSxS) or a 
            numpy ndarray of dytpe=object with 1 dimenion (1xA), each 
            element containing a numpy ndarray (SxS) or scipy sparse matrix.
        
        reward = reward matrix
            R could be a numpy ndarray with 3 dimensions (AxSxS) or numpy
            ndarray of dtype=object with 1 dimension (1xA), each element
            containing a sparse matrix (SxS). R also could be a numpy 
            ndarray with 2 dimensions (SxA) possibly sparse.
        
        discount = discount rate in ]0; 1]
            Beware to check conditions of convergence for discount = 1.
        
        epsilon = epsilon-optimal policy search
            Greater than 0, optional (default: 0.01).
        
        max_iter = maximum number of iterations to be done.
            greater than 0, optional (default: computed)
        
        initial_value = starting value function.
            optional (default: zeros(S,1)).
        
        Evaluation
        ----------
        value(S) = value function.
        
        policy(S) = epsilon-optimal policy.
        
        iter = number of done iterations.
        
        time = used CPU time.
        
        Notes
        -----
Steven Cordwell's avatar
Steven Cordwell committed
622
623
624
625
626
        In verbose mode, at each iteration, displays the variation of V
        and the condition which stopped iterations: epsilon-optimum policy found
        or maximum number of iterations reached.
        """
        
Steven Cordwell's avatar
Steven Cordwell committed
627
        MDP.__init__()
Steven Cordwell's avatar
Steven Cordwell committed
628
        
Steven Cordwell's avatar
Steven Cordwell committed
629
        self.check(transitions, reward)
Steven Cordwell's avatar
Steven Cordwell committed
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655

        self.computePR(transitions, reward)
        
        # initialization of optional arguments
        if (initial_value == 0):
            self.value = matrix(zeros((self.S, 1)))
        else:
            if (initial_value.size != self.S):
                raise TypeError("The initial value must be length S")
            
            self.value = matrix(initial_value)
        
        self.discount = discount
        if (discount < 1):
            # compute a bound for the number of iterations
            #self.max_iter = self.boundIter(epsilon)
            self.max_iter = 5000
            # computation of threshold of variation for V for an epsilon-optimal policy
            self.thresh = epsilon * (1 - self.discount) / self.discount
        else: # discount == 1
            # bound for the number of iterations
            self.max_iter = max_iter
            # threshold of variation for V for an epsilon-optimal policy
            self.thresh = epsilon 
        
        self.itr = 0
Steven Cordwell's avatar
Steven Cordwell committed
656
657
        
        self.time = None
Steven Cordwell's avatar
Steven Cordwell committed
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
    
    def boundIter(self, epsilon):
        """Computes a bound for the number of iterations for the value iteration
        algorithm to find an epsilon-optimal policy with use of span for the 
        stopping criterion
        
        Arguments --------------------------------------------------------------
        Let S = number of states, A = number of actions
            epsilon   = |V - V*| < epsilon,  upper than 0,
                optional (default : 0.01)
        Evaluation -------------------------------------------------------------
            max_iter  = bound of the number of iterations for the value 
            iteration algorithm to find an epsilon-optimal policy with use of
            span for the stopping criterion
            cpu_time  = used CPU time
        """
        # See Markov Decision Processes, M. L. Puterman, 
        # Wiley-Interscience Publication, 1994 
        # p 202, Theorem 6.6.6
        # k =    max     [1 - S min[ P(j|s,a), p(j|s',a')] ]
        #     s,a,s',a'       j
        k = 0
        h = zeros(self.S)
        
        for ss in range(self.S):
            PP = zeros((self.S, self.A))
            for aa in range(self.A):
                PP[:, aa] = self.P[aa][:, ss]
            # the function "min()" without any arguments finds the
            # minimum of the entire array.
            h[ss] = PP.min()
        
        k = 1 - h.sum()
        V1 = self.bellmanOperator(self.value)
        # p 201, Proposition 6.6.5
        max_iter = log( (epsilon * (1 - self.discount) / self.discount) / self.span(V1 - self.value) ) / log(self.discount * k)
        return ceil(max_iter)
    
    def iterate(self):
        """
        """
        self.time = time()
        done = False
        while not done:
            self.itr = self.itr + 1
            
            Vprev = self.value
            
            # Bellman Operator: updates "self.value" and "self.policy"
Steven Cordwell's avatar
Steven Cordwell committed
707
            self.bellmanOperator()
Steven Cordwell's avatar
Steven Cordwell committed
708
            
Steven Cordwell's avatar
Steven Cordwell committed
709
710
            # The values, based on Q. For the function "max()": the option
            # "axis" means the axis along which to operate. In this case it
Steven Cordwell's avatar
Steven Cordwell committed
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
            # finds the maximum of the the rows. (Operates along the columns?)
            variation = self.span(self.value - Vprev)
            
            if self.verbose:
                print("      %s         %s" % (self.itr, variation))
            
            if variation < self.thresh:
                done = True
                if self.verbose:
                    print("...iterations stopped, epsilon-optimal policy found")
            elif (self.itr == self.max_iter):
                done = True 
                if self.verbose:
                    print("...iterations stopped by maximum number of iteration condition")

        self.value = array(self.value).reshape(self.S)
        self.policy = array(self.policy).reshape(self.S)

        self.time = time() - self.time

Steven Cordwell's avatar
Steven Cordwell committed
731
732
733
734
class ValueIterationGS(MDP):
    """Resolution of discounted MDP with value iteration Gauss-Seidel algorithm.
    """
    pass