pyspark.sql.DataFrameStatFunctions.approxQuantile

DataFrameStatFunctions.approxQuantile(col, probabilities, relativeError)

Calculates the approximate quantiles of numerical columns of a DataFrame.

The result of this algorithm has the following deterministic bound: If the DataFrame has N elements and if we request the quantile at probability p up to error err, then the algorithm will return a sample x from the DataFrame so that the exact rank of x is close to (p * N). More precisely,

floor((p - err) * N) <= rank(x) <= ceil((p + err) * N).

This method implements a variation of the Greenwald-Khanna algorithm (with some speed optimizations). The algorithm was first present in [[https://doi.org/10.1145/375663.375670 Space-efficient Online Computation of Quantile Summaries]] by Greenwald and Khanna.

New in version 2.0.0.

Changed in version 3.4.0: Supports Spark Connect.

Parameters

col: str, tuple or list

Can be a single column name, or a list of names for multiple columns.

Changed in version 2.2.0: Added support for multiple columns.

probabilitieslist or tuple of floats

a list of quantile probabilities Each number must be a float in the range [0, 1]. For example 0.0 is the minimum, 0.5 is the median, 1.0 is the maximum.

relativeErrorfloat

The relative target precision to achieve (>= 0). If set to zero, the exact quantiles are computed, which could be very expensive. Note that values greater than 1 are accepted but gives the same result as 1.

Returns

list

the approximate quantiles at the given probabilities.

• If the input col is a string, the output is a list of floats.

• If the input col is a list or tuple of strings, the output is also a

  list, but each element in it is a list of floats, i.e., the output is a list of list of floats.

Notes

Null values will be ignored in numerical columns before calculation. For columns only containing null values, an empty list is returned.

Examples

Example 1: Calculating quantiles for a single column

>>> data = [(1,), (2,), (3,), (4,), (5,)]
>>> df = spark.createDataFrame(data, ["values"])
>>> quantiles = df.approxQuantile("values", [0.0, 0.5, 1.0], 0.05)
>>> quantiles
[1.0, 3.0, 5.0]

Example 2: Calculating quantiles for multiple columns

>>> data = [(1, 10), (2, 20), (3, 30), (4, 40), (5, 50)]
>>> df = spark.createDataFrame(data, ["col1", "col2"])
>>> quantiles = df.approxQuantile(["col1", "col2"], [0.0, 0.5, 1.0], 0.05)
>>> quantiles
[[1.0, 3.0, 5.0], [10.0, 30.0, 50.0]]

Example 3: Handling null values

>>> data = [(1,), (None,), (3,), (4,), (None,)]
>>> df = spark.createDataFrame(data, ["values"])
>>> quantiles = df.approxQuantile("values", [0.0, 0.5, 1.0], 0.05)
>>> quantiles
[1.0, 3.0, 4.0]

Example 4: Calculating quantiles with low precision

>>> data = [(1,), (2,), (3,), (4,), (5,)]
>>> df = spark.createDataFrame(data, ["values"])
>>> quantiles = df.approxQuantile("values", [0.0, 0.2, 1.0], 0.1)
>>> quantiles
[1.0, 1.0, 5.0]