more testing and cleanup

Author: amogorkon

Showcase.ipynb

@@ -13,102 +13,91 @@
    ]
   },
   {
-   "cell_type": "code",
-   "execution_count": 9,
-   "metadata": {
-    "collapsed": false
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "1"
-      ]
-     },
-     "execution_count": 9,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "cell_type": "markdown",
+   "metadata": {},
    "source": [
-    "from fuzzy.functions import alpha, singleton, constant\n",
-    "\n",
+    "## Usage\n",
     "\n",
-    "a = alpha(0.2, 0.8)\n",
-    "s = singleton(5)\n",
-    "\n"
+    "1. Define a Domain and specify the measurable range & resolution.\n",
+    "2. Define some fuzzysets (\"adjectives\") as attributes that describe ranges within\n",
+    "the domain using Set(). Each set is defined by its membership function func(REAL) -> [0,1]\n",
+    "3. You can combine fuzzysets of the same domain using fuzzy logic operators (~, &, |, +, *) to\n",
+    "define derived fuzzysets.\n",
+    "4. Now suppose you have a value x. Call the fuzzyset(x) directly to produce a\n",
+    "membership value or call the domain with it to get a dictionary of memberships.\n",
+    "5. To really put these to work, you will want to map membership values of one or\n",
+    "more domains onto another domain, producing \"real world\" values again that can\n",
+    "be used to control things. This is called inference (combining domains) and\n",
+    "defuzzification (turning membership values back into measurables)."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "metadata": {
     "collapsed": false
    },
-   "outputs": [
-    {
-     "ename": "TypeError",
-     "evalue": "'>=' not supported between instances of 'function' and 'float'",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
-      "\u001b[0;32m<ipython-input-3-7ecf006a5990>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mnumbers\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mDomain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"numbers\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlow\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhigh\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m20\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0mclose_to_10\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mSet\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnumbers\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0.2\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0.8\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtriangular\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m20\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m \u001b[0mclose_to_10\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m/media/sf_Google_Drive/jupyter/fuzzy/fuzzy/functions.py\u001b[0m in \u001b[0;36mf_value\u001b[0;34m(x)\u001b[0m\n\u001b[1;32m     90\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     91\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mf_value\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 92\u001b[0;31m         \u001b[0;32mif\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m>=\u001b[0m \u001b[0mceiling\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     93\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mceiling\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     94\u001b[0m         \u001b[0;32melif\u001b[0m \u001b[0mx\u001b[0m \u001b[0;34m<=\u001b[0m \u001b[0mfloor\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;31mTypeError\u001b[0m: '>=' not supported between instances of 'function' and 'float'"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
+    "%matplotlib inline\n",
     "from fuzzy.classes import Domain, Set\n",
     "from fuzzy.functions import alpha, triangular\n",
+    "from fuzzy.hedges import plus, minus\n",
     "\n",
-    "numbers = Domain(\"numbers\", low=0, high=20)\n",
-    "close_to_10 = Set(numbers, alpha(0.2,0.8)(triangular(0, 20)))\n",
-    "close_to_10.plot()"
+    "numbers = Domain(\"numbers\", low=0, high=20, res=0.1)\n",
+    "close_to_10 = Set(numbers, alpha(0.2, 0.8, \n",
+    "                            triangular(0, 20)\n",
+    "                            ))\n",
+    "close_to_5 = Set(numbers, triangular(1,10))\n",
+    "foo = plus(minus(close_to_5))\n",
+    "bar = close_to_10\n",
+    "bar.plot()\n",
+    "foo.plot()\n",
+    "baz = foo + bar\n",
+    "baz.plot()"
    ]
   },
   {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "collapsed": false
-   },
-   "outputs": [],
+   "cell_type": "markdown",
+   "metadata": {},
    "source": [
-    "%matplotlib inline\n",
-    "from fuzzy.classes import Domain, Set, Rule\n",
-    "from fuzzy.functions import S, R\n",
+    "## Rules\n",
     "\n",
-    "zeit = Domain(\"time\", low=0, high=24, res=0.1)\n",
-    "morgen = Set(zeit, S(0, 8))\n",
-    "abend = Set(zeit, R(18, 24))\n",
-    "mittag = ~morgen & ~abend\n",
-    "\n",
-    "morgen.plot()\n",
-    "abend.plot()\n",
-    "mittag.plot()\n"
+    "Rules aka Fuzzy Inference is done when domains and sets are defined to combine sets of different domains. "
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 7,
    "metadata": {
     "collapsed": false
    },
-   "outputs": [],
+   "outputs": [
+    {
+     "ename": "AttributeError",
+     "evalue": "'NotImplementedType' object has no attribute 'copy'",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
+      "\u001b[0;32m<ipython-input-7-4a94e0f36941>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0md1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtemp\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m32\u001b[0m\u001b[0;34m)\u001b[0m   \u001b[0;31m# {'temperature.hot': 1}\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[0md2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m5\u001b[0m\u001b[0;34m)\u001b[0m    \u001b[0;31m# {'distance.close': 0}\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 10\u001b[0;31m \u001b[0md\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0md1\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcopy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m   \u001b[0;31m# need to merge the results of the Domains\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     11\u001b[0m \u001b[0md\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mupdate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0md2\u001b[0m\u001b[0;34m)\u001b[0m    \u001b[0;31m# for py3.5: https://www.python.org/dev/peps/pep-0448/\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     12\u001b[0m \u001b[0mr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0md\u001b[0m\u001b[0;34m)\u001b[0m    \u001b[0;31m# min(1, 0)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
+      "\u001b[0;31mAttributeError\u001b[0m: 'NotImplementedType' object has no attribute 'copy'"
+     ]
+    }
+   ],
    "source": [
     "from fuzzy.classes import Domain, Set, Rule\n",
-    "from fuzzy.functions import S, R\n",
-    "\n",
-    "dist = Domain(\"distance\", low=0, high=1000, res=0.1)\n",
-    "near = Set(dist, S(0, 100))\n",
-    "\n",
-    "standing = Domain(\"standing\", low=0, high=100)\n",
-    "friendly = Set(standing, R(0, 100))\n",
-    "\n",
-    "reaction = Domain(\"reaction\", low=0, high=10)\n",
-    "aggro = IF(near & ~friendly, reaction)"
+    "    \n",
+    "temp = Domain(\"temperature\", 0, 100)\n",
+    "temp.hot = Set(temp, lambda x: 1)\n",
+    "dist = Domain(\"distance\", 0, 300)\n",
+    "dist.close = Set(dist, lambda x: 0)\n",
+    "r = Rule(min, [\"distance.close\", \"temperature.hot\"])\n",
+    "d1 = temp(32)   # {'temperature.hot': 1}\n",
+    "d2 = dist(5)    # {'distance.close': 0}\n",
+    "d = d1.copy()   # need to merge the results of the Domains\n",
+    "d.update(d2)    # for py3.5: https://www.python.org/dev/peps/pep-0448/\n",
+    "r(d)    # min(1, 0)"
    ]
   },
   {

TODO-fuzzy

@@ -1,5 +1,15 @@
-do: build a testcase suite
+do: build a #testcase suite
 ---
-do: review functions
+do: MUST review #functions
 ---
-do: Set.plot should always plot between 0 and 1
+do: #Set.plot SHOULD always plot between 0 and 1
+---
+do: #Set.plot SHOULD always plot the full set range
+---
+do: MUST fix #Domain._sets (should be either set OR dict)
+until: 2017-3-1
+---
+do: check tests
+every: day
+---
+do: SHOULD find examples for truth functions

fuzzy/classes.py

@@ -1,27 +1,9 @@
 
-"""
-Fuzzy Logic for Python 3.
-
-Usage:
-1) Define a Domain and specify the measurable range & resolution.
-2) Define some fuzzysets ("adjectives") as attributes that describe ranges within
-the domain using Set().
-3) You can combine fuzzysets using fuzzy logic operators (~, &, |, +, *) to
-define derived fuzzysets.
-4) Now suppose you have a value x. Call the fuzzyset(x) directly to produce a
-membership value or
-call the domain with it to get a dictionary of memberships.
-5) To really put these to work, you will want to map membership values of one or
-more domains onto another domain, producing "real world" values again that can
-be used to control things. This is called inference (combining domains) and
-defuzzification (turning membership values back into measurables).
-"""
-
 import matplotlib.pyplot as plt
 from numpy import arange
 
-from functions import inv
-from combinators import MAX, MIN, product, bounded_sum
+from fuzzy.functions import inv
+from fuzzy.combinators import MAX, MIN, product, bounded_sum
 
 
 
@@ -125,44 +107,43 @@ class Set:
 
     Sets that are returned from one of the operations are 'derived sets' or
     'Superfuzzysets' according to Zadeh.
+    
+    Note that most checks are merely assertions that can be optimized away.
+    DO NOT RELY on these checks and use tests to make sure that only valid calls are made.
     """
     def __init__(self, domain:Domain, func:callable, ops:dict=None):
         self.func = func
         assert self.func is not None
         self.ops = ops
-        if not isinstance(domain, Domain):
-            raise AttributeError
+        assert isinstance(domain, Domain), "Must be used with a Domain."
         self.domain = domain
         self.domain._sets.add(self)
 
     def __call__(self, x):
+        assert self.domain.low <= x <= self.domain.high, "value outside domain"
         return self.func(x)
 
     def __invert__(self):
         return Set(self.domain, inv(self.func))
 
     def __and__(self, other):
-        if self.domain is not other.domain:
-            raise UserWarning
+        assert self.domain is other.domain, "Cannot combine sets of different domains."
         return Set(self.domain, MIN(self.func, other.func))
 
     def __or__(self, other):
-        if self.domain is not other.domain:
-            raise UserWarning
+        assert self.domain is other.domain, "Cannot combine sets of different domains."
         return Set(self.domain, MAX(self.func, other.func))
 
     def __mul__(self, other):
-        if self.domain is not other.domain:
-            raise UserWarning
+        assert self.domain is other.domain, "Cannot combine sets of different domains."
         return Set(self.domain, product(self.func, other.func))
 
-    def __sum__(self, other):
-        if self.domain is not other.domain:
-            raise UserWarning
+    def __add__(self, other):
+        assert self.domain is other.domain, "Cannot combine sets of different domains."
         return Set(self.domain, bounded_sum(self.func, other.func))
 
     def __pow__(self, power):
-        """pow is used with lingual hedges as defined in fuzzy.functions"""
+        """pow is used with hedges"""
         return Set(self.domain, lambda x: pow(self.func(x), power))
 
     def plot(self, low=None, high=None, res=None):
@@ -175,8 +156,6 @@ def plot(self, low=None, high=None, res=None):
         res = self.domain.res if res is None else res
         R = arange(low, high, res)
         V = [self.func(x) for x in R]
-        #print(R)
-        #print(V)
         plt.plot(R, V)
 
 

fuzzy/functions.py

@@ -399,8 +399,8 @@ def bounded_sigmoid(low, high):
     >>> round(f(-100000), 2)
     0.0
     """
-    if high < low:
-        raise ValueError('high must not be less than low.')
+    assert low < high, 'high must not be less than low'
+    
     k = -(4. * log(3)) / (low - high)
     o = 9 * exp(low * k)
 

fuzzy/hedges.py

@@ -7,7 +7,7 @@
 These work with sets only. Meta-functionality is deprecated.
 """
 
-from classes import Set
+from fuzzy.classes import Set
 from warnings import warn
 
 def very(g):

fuzzy/truth.py

@@ -4,6 +4,8 @@
 TRUTH QUALIFICATION
 -------------------
 Functions that transform a given membership value to a truth value.
+
+How this can be useful? Well.. 
 """
 
 

test_fuzzy_functionality.py

@@ -1,3 +1,4 @@
+
 """
 Functional test of the fuzzylogic lib 'fuzzy'.
 """
@@ -35,4 +36,4 @@ def test_temp(self):
 
 
 if __name__ == '__main__':
-    unittest.main()
+    unittest.main()

test_fuzzy_units.py

@@ -1,22 +1,19 @@
 
 from hypothesis import given, strategies as st, assume
+from math import isclose
 
 from fuzzy import functions as fun
-
-
+    
 @given(st.floats(allow_nan=False, allow_infinity=False))
 def test_noop(x):
     f = fun.noop()
     assert f(x) == x
 
 @given(st.floats(allow_nan=False, allow_infinity=False))
-def test_invert(x):
-    f = fun.inv()
-    assert f(x) == 1 - x
-    
-    n = fun.noop()
-    f = fun.inv(n)
-    assert f(x) == 1 - x
+def test_inv(x):
+    assume(0 <= x <= 1)
+    f = fun.inv(fun.noop())
+    assert isclose(f(f(x)), x, abs_tol=1e-16)
 
 @given(st.floats(allow_nan=False, allow_infinity=False), 
        st.floats(allow_nan=False, allow_infinity=False))
@@ -30,8 +27,7 @@ def test_constant(c, r):
        st.floats(allow_nan=False))
 def test_alpha(lower, upper, x):
     assume(lower < upper)
-        
-    f = fun.alpha(lower, upper)
+    f = fun.alpha(lower, upper, fun.noop())
     if x <= lower:
         assert f(x) == lower
     elif x >= upper: