pyspark.RDD.countApproxDistinct

RDD.countApproxDistinct(relativeSD: float = 0.05) → int

Return approximate number of distinct elements in the RDD.

Parameters
relativeSDfloat, optional

Relative accuracy. Smaller values create counters that require more space. It must be greater than 0.000017.

Notes

The algorithm used is based on streamlib’s implementation of “HyperLogLog in Practice: Algorithmic Engineering of a State of The Art Cardinality Estimation Algorithm”, available here.

Examples

>>> n = sc.parallelize(range(1000)).map(str).countApproxDistinct()
>>> 900 < n < 1100
True
>>> n = sc.parallelize([i % 20 for i in range(1000)]).countApproxDistinct()
>>> 16 < n < 24
True