`ranked.py`

Note that this is obsolete. The elections module should be used instead.

The sample Python session below shows off most of the functions of the ranked module. Hopefully, the comments in the code explain the rest. Note that parseBLT('foo.blt’) can take several seconds; however, the first time a .blt file is parsed, a cached version with the extension .pickle is written which makes subsequent calls to parseBLT('foo.blt’) significantly faster.

>>> import ranked
>>> root, numBallots, ranks, candidates, description = ranked.parseBLT('2010-Oakland-Mayor.blt')
>>> description
'Mayor - Oakland (RCV)'
>>> winner, counts, elimination = ranked.irv(root, rules=ranked.SFRCVRules)
>>> candidates[winner]
'JEAN QUAN'
>>> for c in sorted(candidates):
...     print '%30s:' % candidates[c], counts[c]
...
                    DON PERATA: [40342, 45465, 51872]
               TERENCE CANDELL: [2315, 0, 0]
                  GREG HARLAND: [966, 0, 0]
                   DON MACLEAY: [1630, 0, 0]
                     JEAN QUAN: [29266, 35033, 53897]
                 ARNOLD FIELDS: [733, 0, 0]
                     JOE TUMAN: [14347, 0, 0]
                  MARCIE HODGE: [2994, 0, 0]
 LARRY LIONEL ''LL'' YOUNG JR.: [933, 0, 0]
                REBECCA KAPLAN: [25813, 32719, 0]
                      Write-In: [268, 0, 0]
>>> ranked.irvUB(root)
2026.0
>>> ranked.irvLB(root)[0]
2025
>>> m = ranked.buildCondorcet(root)
>>> candidates[ranked.condorcetWinner(m)]
'JEAN QUAN'
>>> ranked.condorcetMargin(m)
2025