René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Exceptions from `round` implementation","articleBody":"[Prompted by [this StackOverflow question](https://stackoverflow.com/q/48360030/270986)]\r\n\r\nThere are a couple of issues with the current implementation of `round` from `builtins`:\r\n\r\n- it doesn't support negative second argument:\r\n\r\n```\r\n\u003e\u003e\u003e from builtins import round\r\n\u003e\u003e\u003e round(314, -2)\r\nTraceback (most recent call last):\r\n  File \"\u003cstdin\u003e\", line 1, in \u003cmodule\u003e\r\n  File \"/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/future/builtins/newround.py\", line 33, in newround\r\n    raise NotImplementedError('negative ndigits not supported yet')\r\nNotImplementedError: negative ndigits not supported yet\r\n```\r\n\r\n- it can raise `decimal.InvalidOperation`:\r\n\r\n```\r\n\u003e\u003e\u003e round(1e50, 2)\r\nTraceback (most recent call last):\r\n  File \"\u003cstdin\u003e\", line 1, in \u003cmodule\u003e\r\n  File \"/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/future/builtins/newround.py\", line 43, in newround\r\n    rounding=ROUND_HALF_EVEN)\r\n  File \"/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/decimal.py\", line 2468, in quantize\r\n    'quantize result has too many digits for current context')\r\n  File \"/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/decimal.py\", line 3872, in _raise_error\r\n    raise error(explanation)\r\ndecimal.InvalidOperation: quantize result has too many digits for current context\r\n```\r\n\r\nThe first restriction seems unnecessary. The second can be overcome by using a dedicated `decimal` context with sufficient precision. For rounding a float `x` to `n` decimal places (with `n` possibly negative), I'd suggest something like the following:\r\n\r\n```\r\nfrom decimal import Context, Decimal\r\nc = Context(prec=800)  # sufficient precision that the `quantize` result is representable\r\nquantum = Decimal(\"1e{}\".format(-n))  # should work for both n positive and negative\r\nrounded = float(c.quantize(Decimal(x), quantum))\r\n```\r\n\r\n(The bound of 800 here is a bit crude. The maximum number of significant digits needed to represent any IEEE 754 binary64 float exactly in decimal is 767.)","author":{"url":"https://github.com/mdickinson","@type":"Person","name":"mdickinson"},"datePublished":"2018-01-20T19:46:04.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/322/python-future/issues/322"}