diff options
Diffstat (limited to 'python/pyspark/statcounter.py')
-rw-r--r-- | python/pyspark/statcounter.py | 22 |
1 files changed, 22 insertions, 0 deletions
diff --git a/python/pyspark/statcounter.py b/python/pyspark/statcounter.py index 0fee3b2096..03ea0b6d33 100644 --- a/python/pyspark/statcounter.py +++ b/python/pyspark/statcounter.py @@ -131,6 +131,28 @@ class StatCounter(object): def sampleStdev(self): return sqrt(self.sampleVariance()) + def asDict(self, sample=False): + """Returns the :class:`StatCounter` members as a ``dict``. + + >>> sc.parallelize([1., 2., 3., 4.]).stats().asDict() + {'count': 4L, + 'max': 4.0, + 'mean': 2.5, + 'min': 1.0, + 'stdev': 1.2909944487358056, + 'sum': 10.0, + 'variance': 1.6666666666666667} + """ + return { + 'count': self.count(), + 'mean': self.mean(), + 'sum': self.sum(), + 'min': self.min(), + 'max': self.max(), + 'stdev': self.stdev() if sample else self.sampleStdev(), + 'variance': self.variance() if sample else self.sampleVariance() + } + def __repr__(self): return ("(count: %s, mean: %s, stdev: %s, max: %s, min: %s)" % (self.count(), self.mean(), self.stdev(), self.max(), self.min())) |