Skip to main content
Version: Java (Groovy)

AggPct

AggPct returns an aggregator that computes the designated percentile of values, within an aggregation group, for each input column.

Syntax​

AggPct(percentile, average, pairs...)
AggPct(percentile, pairs...)
AggPct(inputColumn, average, percentileOutputs...)
AggPct(inputColumn, percentileOutputs...)

Parameters​

ParameterTypeDescription
percentiledouble

The percentile to use for all component aggregations.

pairsString...

The input/output column name pairs.

  • "X" will output the designated percentile value in the X column for each group.
  • "Y = X" will output the designated percentile value in the X column for each group and rename it to Y.
  • "X, A = B" will output the designated percentile value in the X column for each group and the designated percentile value in the B value column renaming it to A.
averageboolean

Whether to average the two middle values for even-sized result sets of numeric types. See AggSpecMedian.

percentileOutputsPercentileOutput...

The percentile/output column name pairs for the component aggregations. See PctOut(double, String)

caution

If an aggregation does not rename the resulting column, the aggregation column will appear in the output table, not the input column. If multiple aggregations on the same column do not rename the resulting columns, an error will result, because the aggregations are trying to create multiple columns with the same name. For example, in table.aggBy([agg.AggSum(“X”), agg.AggAvg(“X”)]), both the sum and the average aggregators produce column X, which results in an error.

Returns​

An aggregator that computes the designated percentile value, within an aggregation group, for each input column.

Examples​

In this example, AggPct returns the 68th percentile value Number as grouped by X.

import static io.deephaven.api.agg.Aggregation.AggPct

source = newTable(
    stringCol("X", "A", "B", "A", "C", "B", "A", "B", "B", "C"),
    stringCol("Y", "M", "N", "O", "N", "P", "M", "O", "P", "M"),
    intCol("Number", 55, 76, 20, 130, 230, 50, 73, 137, 214),
)

result = source.aggBy([AggPct(0.68, "PctNumber = Number")], "X")

In this example, AggPct returns the 68th percentile value Number and the 99th percentile value Number as grouped by X.

import static io.deephaven.api.agg.Aggregation.AggPct
import static io.deephaven.api.agg.Aggregation.AggPct

source = newTable(
    stringCol("X", "A", "B", "A", "C", "B", "A", "B", "B", "C"),
    stringCol("Y", "M", "N", "O", "N", "P", "M", "O", "P", "M"),
    intCol("Number", 55, 76, 20, 130, 230, 50, 73, 137, 214),
)

result = source.aggBy([AggPct(0.68, "Pct68Number = Number"),AggPct(0.99, "Pct99Number = Number")], "X")

In this example, AggPct returns the 97th percentile value Number, and AggMed returns the median Number, as grouped by X.

import static io.deephaven.api.agg.Aggregation.AggPct
import static io.deephaven.api.agg.Aggregation.AggMed

source = newTable(
    stringCol("X", "A", "B", "A", "C", "B", "A", "B", "B", "C"),
    stringCol("Y", "M", "P", "O", "N", "P", "M", "O", "P", "N"),
    intCol("Number", 55, 76, 20, 130, 230, 50, 73, 137, 214),
)

result = source.aggBy([AggPct(0.97, "Pct97Number = Number"),AggMed("MedNumber = Number")], "X")