[Python-checkins] r79609 - in python/trunk: Doc/library/decimal.rst Lib/decimal.py Lib/test/test_decimal.py

raymond.hettinger python-checkins at python.org
Fri Apr 2 20:39:24 CEST 2010


Author: raymond.hettinger
Date: Fri Apr  2 20:39:24 2010
New Revision: 79609

Log:
Issue 8257: Decimal constructor to accept float.

Modified:
   python/trunk/Doc/library/decimal.rst
   python/trunk/Lib/decimal.py
   python/trunk/Lib/test/test_decimal.py

Modified: python/trunk/Doc/library/decimal.rst
==============================================================================
--- python/trunk/Doc/library/decimal.rst	(original)
+++ python/trunk/Doc/library/decimal.rst	Fri Apr  2 20:39:24 2010
@@ -314,7 +314,7 @@
 
    Construct a new :class:`Decimal` object based from *value*.
 
-   *value* can be an integer, string, tuple, or another :class:`Decimal`
+   *value* can be an integer, string, tuple, :class:`float`, or another :class:`Decimal`
    object. If no *value* is given, returns ``Decimal('0')``.  If *value* is a
    string, it should conform to the decimal numeric string syntax after leading
    and trailing whitespace characters are removed::
@@ -341,6 +341,11 @@
    digits, and an integer exponent. For example, ``Decimal((0, (1, 4, 1, 4), -3))``
    returns ``Decimal('1.414')``.
 
+   If *value* is a :class:`float`, the binary floating point value is losslessly
+   converted to its exact decimal equivalent.  This conversion can often require
+   upto 53 digits of precision.  For example, ``Decimal(float('1.1'))`` converts
+   to ``Decimal('1.100000000000000088817841970012523233890533447265625')``.
+
    The *context* precision does not affect how many digits are stored. That is
    determined exclusively by the number of digits in *value*. For example,
    ``Decimal('3.00000')`` records all five zeros even if the context precision is
@@ -1948,17 +1953,16 @@
         '''
         return d.quantize(Decimal(1)) if d == d.to_integral() else d.normalize()
 
-Q. Is there a way to convert a regular float to a Decimal?
+Q. Is there a way to convert a regular float to a :class:`Decimal`?
 
-A. Yes, the classmethod :meth:`from_float` makes an exact conversion.
+A. Yes, all binary floating point numbers can be exactly expressed as a
+Decimal though an exact conversion may take more precision than intuition would
+suggest:
 
-The regular decimal constructor does not do this by default because there is
-some question about whether it is advisable to mix binary and decimal floating
-point. Also, its use requires some care to avoid the representation issues
-associated with binary floating point:
+.. doctest::
 
-   >>> Decimal.from_float(1.1)
-   Decimal('1.100000000000000088817841970012523233890533447265625')
+    >>> Decimal(math.pi)
+    Decimal('3.141592653589793115997963468544185161590576171875')
 
 Q. Within a complex calculation, how can I make sure that I haven't gotten a
 spurious result because of insufficient precision or rounding anomalies.

Modified: python/trunk/Lib/decimal.py
==============================================================================
--- python/trunk/Lib/decimal.py	(original)
+++ python/trunk/Lib/decimal.py	Fri Apr  2 20:39:24 2010
@@ -648,8 +648,12 @@
             return self
 
         if isinstance(value, float):
-            raise TypeError("Cannot convert float in Decimal constructor. "
-                            "Use from_float class method.")
+            value = Decimal.from_float(value)
+            self._exp  = value._exp
+            self._sign = value._sign
+            self._int  = value._int
+            self._is_special  = value._is_special
+            return self
 
         raise TypeError("Cannot convert %r to Decimal" % value)
 

Modified: python/trunk/Lib/test/test_decimal.py
==============================================================================
--- python/trunk/Lib/test/test_decimal.py	(original)
+++ python/trunk/Lib/test/test_decimal.py	Fri Apr  2 20:39:24 2010
@@ -54,6 +54,11 @@
         )
     setcontext(DefaultTestContext)
 
+# decorator for skipping tests on non-IEEE 754 platforms
+requires_IEEE_754 = unittest.skipUnless(
+    float.__getformat__("double").startswith("IEEE"),
+    "test requires IEEE 754 doubles")
+
 TESTDATADIR = 'decimaltestdata'
 if __name__ == '__main__':
     file = sys.argv[0]
@@ -509,6 +514,27 @@
         self.assertEqual(str(e), '0')
         self.assertNotEqual(id(d), id(e))
 
+    @requires_IEEE_754
+    def test_explicit_from_float(self):
+        r = Decimal(0.1)
+        self.assertEqual(type(r), Decimal)
+        self.assertEqual(str(r),
+                '0.1000000000000000055511151231257827021181583404541015625')
+        self.assertTrue(Decimal(float('nan')).is_qnan())
+        self.assertTrue(Decimal(float('inf')).is_infinite())
+        self.assertTrue(Decimal(float('-inf')).is_infinite())
+        self.assertEqual(str(Decimal(float('nan'))),
+                         str(Decimal('NaN')))
+        self.assertEqual(str(Decimal(float('inf'))),
+                         str(Decimal('Infinity')))
+        self.assertEqual(str(Decimal(float('-inf'))),
+                         str(Decimal('-Infinity')))
+        self.assertEqual(str(Decimal(float('-0.0'))),
+                         str(Decimal('-0')))
+        for i in range(200):
+            x = random.expovariate(0.01) * (random.random() * 2.0 - 1.0)
+            self.assertEqual(x, float(Decimal(x))) # roundtrip
+
     def test_explicit_context_create_decimal(self):
 
         nc = copy.copy(getcontext())


More information about the Python-checkins mailing list