renaming pyrlap to msdm

MANIFEST.in

@@ -1 +1 @@
-include README.rst
+include README.md

README.md

@@ -0,0 +1,21 @@
+# Models of Sequential Decision Making
+
+Implementations of basic rl and planning algorithms and domains
+mainly for cog sci research.
+
+
+## Installation
+
+### Installing the package in edit mode
+
+After downloading, go into the folder and install the package locally
+(with a symlink so its updated as source file changes are made):
+
+```
+$ pip install -e .
+```
+
+It is recommended to use a virtual environment.
+
+Related libraries:
+- [BURLAP](https://github.com/jmacglashan/burlap)

demos/Distributions.ipynb

@@ -9,14 +9,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
     "import pandas as pd\n",
     "\n",
-    "from pyrlap.core.distributions import DiscreteFactorTable as Pr\n",
-    "from pyrlap.core.assignment import AssignmentMap as Dict\n",
+    "from msdm.core.distributions import DiscreteFactorTable as Pr\n",
+    "from msdm.core.assignment import AssignmentMap as Dict\n",
     "\n",
     "def toDF(p):\n",
     "    df = pd.DataFrame(p.support)\n",
@@ -58,7 +58,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -129,7 +129,7 @@
        "3  1  1  0.05 -2.995732"
       ]
      },
-     "execution_count": 3,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -144,7 +144,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -196,7 +196,7 @@
        "1  2  0.285714 -1.252763"
       ]
      },
-     "execution_count": 4,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -211,7 +211,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
@@ -266,7 +266,7 @@
        "1  1  1  0.181818 -1.704748"
       ]
      },
-     "execution_count": 5,
+     "execution_count": 4,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -281,7 +281,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [
     {
@@ -348,7 +348,7 @@
        "2  1  1  1  0.090909 -2.397895"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 5,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -376,7 +376,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -435,7 +435,7 @@
        "2  2  0.826446 -0.190620"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 6,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -450,7 +450,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
@@ -521,7 +521,7 @@
        "3  1  1  0.05 -2.995732"
       ]
      },
-     "execution_count": 8,
+     "execution_count": 7,
      "metadata": {},
      "output_type": "execute_result"
     }

msdm/algorithms/__init__.py

@@ -0,0 +1,4 @@
+from msdm.algorithms.vectorizedvalueiteration import \
+    VectorizedValueIteration
+from msdm.algorithms.laostar import LAOStar
+from msdm.algorithms.lrtdp import LRTDP

pyrlap/algorithms/laostar.py → msdm/algorithms/laostar.py

@@ -4,8 +4,8 @@
 import random
 import numpy as np
 
-from pyrlap.core.assignment import AssignmentMap as Dict
-from pyrlap.core.algorithmclasses import Plans, Result
+from msdm.core.assignment import AssignmentMap as Dict
+from msdm.core.algorithmclasses import Plans, Result
 
 def _hash(x):
     if isinstance(x, dict):

pyrlap/algorithms/lrtdp.py → msdm/algorithms/lrtdp.py

@@ -1,7 +1,7 @@
-from pyrlap.core.problemclasses.mdp import MarkovDecisionProcess
-from pyrlap.core.assignment import DefaultAssignmentMap, \
+from msdm.core.problemclasses.mdp import MarkovDecisionProcess
+from msdm.core.assignment import DefaultAssignmentMap, \
     AssignmentMap
-from pyrlap.core.algorithmclasses import Plans, Result
+from msdm.core.algorithmclasses import Plans, Result
 
 def iter_dist_prob(dist):
     '''

pyrlap/algorithms/vectorizedvalueiteration.py → msdm/algorithms/vectorizedvalueiteration.py

@@ -1,9 +1,9 @@
 from scipy.special import softmax, logsumexp
 import numpy as np
-from pyrlap.core.problemclasses.mdp import TabularPolicy, \
+from msdm.core.problemclasses.mdp import TabularPolicy, \
     TabularMarkovDecisionProcess
-from pyrlap.core.algorithmclasses import Plans, Result
-from pyrlap.core.assignment import AssignmentMap
+from msdm.core.algorithmclasses import Plans, Result
+from msdm.core.assignment import AssignmentMap
 
 class VectorizedValueIteration(Plans):
     def __init__(self,

pyrlap/core/algorithmclasses/__init__.py → msdm/core/algorithmclasses/__init__.py

@@ -1,5 +1,5 @@
 from abc import ABC, abstractmethod
-from pyrlap.core.problemclasses.problemclass import ProblemClass
+from msdm.core.problemclasses.problemclass import ProblemClass
 
 class Algorithm(ABC):
     """Abstract superclass for all algorithms"""

msdm/core/assignment/__init__.py

@@ -0,0 +1,7 @@
+from msdm.core.assignment.assignmentmap import \
+    AssignmentMap, \
+    DefaultAssignmentMap
+from msdm.core.assignment.assignmentset import \
+    AssignmentSet
+from msdm.core.assignment.assignmentcache import \
+    AssignmentCache

msdm/core/distributions/__init__.py

@@ -0,0 +1,4 @@
+from msdm.core.distributions.distributions import Distribution
+from msdm.core.distributions.multinomial import Multinomial
+from msdm.core.distributions.discretefactortable import \
+    DiscreteFactorTable

pyrlap/core/distributions/discretefactortable.py → msdm/core/distributions/discretefactortable.py

@@ -7,8 +7,8 @@
 logger = logging.getLogger(__name__)
 logger.info("Ignoring division by zero errors")
 
-from pyrlap.core.utils.dictutils import dict_merge, dict_match
-from pyrlap.core.distributions.distributions import Distribution
+from msdm.core.utils.dictutils import dict_merge, dict_match
+from msdm.core.distributions.distributions import Distribution
 
 class DiscreteFactorTable(Distribution):
     """

pyrlap/core/distributions/multinomial.py → msdm/core/distributions/multinomial.py

@@ -7,7 +7,7 @@
 logger = logging.getLogger(__name__)
 logger.info("Ignoring division by zero errors")
 
-from pyrlap.core.distributions.distributions import Distribution
+from msdm.core.distributions.distributions import Distribution
 
 class Multinomial(Distribution):
     def __init__(self, support, logits=None, probs=None):

msdm/core/problemclasses/mdp/__init__.py

@@ -0,0 +1,4 @@
+from msdm.core.problemclasses.mdp.mdp import *
+from msdm.core.problemclasses.mdp.tabularmdp import *
+from msdm.core.problemclasses.mdp.factoredmdp import *
+from msdm.core.problemclasses.mdp.policy import *

pyrlap/core/problemclasses/mdp/factoredmdp.py → msdm/core/problemclasses/mdp/factoredmdp.py

@@ -1,4 +1,4 @@
-from pyrlap.core.problemclasses.mdp.mdp import ANDMarkovDecisionProcess, \
+from msdm.core.problemclasses.mdp.mdp import ANDMarkovDecisionProcess, \
     MarkovDecisionProcess
 
 class FactoredMarkovDecisionProcess(MarkovDecisionProcess):

pyrlap/core/problemclasses/mdp/mdp.py → msdm/core/problemclasses/mdp/mdp.py

@@ -1,9 +1,9 @@
 from abc import abstractmethod
 from typing import Iterable
 
-from pyrlap.core.problemclasses.problemclass import ProblemClass
-from pyrlap.core.distributions import Distribution
-from pyrlap.core.assignment.assignmentcache import AssignmentCache
+from msdm.core.problemclasses.problemclass import ProblemClass
+from msdm.core.distributions import Distribution
+from msdm.core.assignment.assignmentcache import AssignmentCache
 
 
 class MarkovDecisionProcess(ProblemClass):

msdm/core/problemclasses/mdp/policy/__init__.py

@@ -0,0 +1,2 @@
+from msdm.core.problemclasses.mdp.policy.tabularpolicy import TabularPolicy
+from msdm.core.problemclasses.mdp.policy.policy import Policy

pyrlap/core/problemclasses/mdp/policy/policy.py → msdm/core/problemclasses/mdp/policy/policy.py

@@ -1,6 +1,6 @@
 from abc import ABC, abstractmethod
-from pyrlap.core.problemclasses.mdp.mdp import MarkovDecisionProcess
-from pyrlap.core.distributions import Distribution
+from msdm.core.problemclasses.mdp.mdp import MarkovDecisionProcess
+from msdm.core.distributions import Distribution
 
 class Policy(ABC):
     @abstractmethod

pyrlap/core/problemclasses/mdp/policy/tabularpolicy.py → msdm/core/problemclasses/mdp/policy/tabularpolicy.py

@@ -1,11 +1,11 @@
 from typing import Mapping
 import numpy as np
 
-from pyrlap.core.problemclasses.mdp.policy.policy import Policy
-from pyrlap.core.problemclasses.mdp.mdp import MarkovDecisionProcess
+from msdm.core.problemclasses.mdp.policy.policy import Policy
+from msdm.core.problemclasses.mdp.mdp import MarkovDecisionProcess
 
-from pyrlap.core.assignment.assignmentmap import AssignmentMap
-from pyrlap.core.distributions import DiscreteFactorTable, Distribution
+from msdm.core.assignment.assignmentmap import AssignmentMap
+from msdm.core.distributions import DiscreteFactorTable, Distribution
 class TabularPolicy(Policy):
     def __init__(self, states, actions, policymatrix=None, policydict=None):
         self._states = states

pyrlap/core/problemclasses/mdp/tabularmdp.py → msdm/core/problemclasses/mdp/tabularmdp.py

@@ -1,9 +1,9 @@
 import json, logging
 import numpy as np
-from pyrlap.core.problemclasses.mdp import ANDMarkovDecisionProcess, \
+from msdm.core.problemclasses.mdp import ANDMarkovDecisionProcess, \
     MarkovDecisionProcess
 
-from pyrlap.core.assignment.assignmentset import AssignmentSet as Set
+from msdm.core.assignment.assignmentset import AssignmentSet as Set
 logger = logging.getLogger(__name__)
 
 class TabularMarkovDecisionProcess(MarkovDecisionProcess):

pyrlap/core/problemclasses/posg/posg.py → msdm/core/problemclasses/posg/posg.py

@@ -1,8 +1,8 @@
 from typing import Mapping, Hashable
 from abc import abstractmethod
 
-from pyrlap.core.problemclasses.problemclass import ProblemClass
-from pyrlap.core.distributions import Distribution
+from msdm.core.problemclasses.problemclass import ProblemClass
+from msdm.core.distributions import Distribution
 
 
 class PartiallyObservableStochasticGame(ProblemClass):

msdm/core/problemclasses/stochasticgame/__init__.py

@@ -0,0 +1,2 @@
+from msdm.core.problemclasses.stochasticgame.stochasticgame \
+    import StochasticGame, ANDStochasticGame

pyrlap/core/problemclasses/stochasticgame/stochasticgame.py → msdm/core/problemclasses/stochasticgame/stochasticgame.py

@@ -1,9 +1,9 @@
 from typing import Mapping, Hashable
 from abc import abstractmethod
 
-from pyrlap.core.problemclasses.problemclass import ProblemClass
-from pyrlap.core.distributions import Distribution
-from pyrlap.core.assignment.assignmentcache import AssignmentCache
+from msdm.core.problemclasses.problemclass import ProblemClass
+from msdm.core.distributions import Distribution
+from msdm.core.assignment.assignmentcache import AssignmentCache
 
 class StochasticGame(ProblemClass):
     """

msdm/domains/__init__.py

@@ -0,0 +1,3 @@
+from msdm.domains.gridworld.mdp import GridWorld
+from msdm.domains.modified.stickyaction import StickyActionMDP
+from msdm.domains.gridworld.stickyactiongw import StickyActionGridWorld

pyrlap/domains/gridgame/gridgame.py → msdm/domains/gridgame/gridgame.py

@@ -3,9 +3,9 @@
 import json, copy
 import numpy as np
 
-from pyrlap.core.utils.gridstringutils import stringToElementArray
-from pyrlap.core.problemclasses.stochasticgame import StochasticGame
-from pyrlap.core.distributions import DiscreteFactorTable as Pr
+from msdm.core.utils.gridstringutils import stringToElementArray
+from msdm.core.problemclasses.stochasticgame import StochasticGame
+from msdm.core.distributions import DiscreteFactorTable as Pr
 
 TERMINALSTATE = {"isTerminal": True}
 

pyrlap/domains/gridworld/mdp.py → msdm/domains/gridworld/mdp.py

@@ -1,13 +1,13 @@
 import matplotlib.pyplot as plt
 import json
 from typing import Iterable
-from pyrlap.core.utils.gridstringutils import  stringToElementArray
+from msdm.core.utils.gridstringutils import  stringToElementArray
 
-from pyrlap.core.problemclasses.mdp import \
+from msdm.core.problemclasses.mdp import \
     TabularMarkovDecisionProcess, \
     ANDMarkovDecisionProcess
-from pyrlap.core.distributions import DiscreteFactorTable
-from pyrlap.core.assignment import \
+from msdm.core.distributions import DiscreteFactorTable
+from msdm.core.assignment import \
     AssignmentMap as Dict, AssignmentSet as Set
 
 def dictToStr(d):
@@ -167,7 +167,7 @@ def plot(self,
         if allElements:
             plotInitStates = True
             plotAbsorbingStates = True
-        from pyrlap.domains.gridworld.plotting import GridWorldPlotter
+        from msdm.domains.gridworld.plotting import GridWorldPlotter
         if featureColors is None:
             featureColors = {
                 'g': 'yellow',