Skip to main content
Version: Python

ByEmaSimple

ByEmaSimple calculates Exponential Moving Averages (EMA) in the Deephaven Query Language.

Syntax​

The constructor for ByEmaSimple is shown below.

ByEmaSimple(
        nullBehavior,
        nanBehavior,
        mode,
        timescale,
        timeUnit,
        type
)

Parameters​

ParameterTypeDescription
nullBehaviorBadDataBehavior
nanBehaviorBadDataBehavior
modeenum
  • TICK uses each new entry of data to determine the EMA window.
  • TIME uses time to determine the EMA window.
timescaledouble

The number of units for the EMA window.

timeUnitenum
  • DAYS - Time unit representing twenty four hours.
  • HOURS - Time unit representing sixty minutes.
  • MICROSECONDS - Time unit representing one thousandth of a millisecond.
  • MILLISECONDS - Time unit representing one thousandth of a second.
  • MINUTES - Time unit representing sixty seconds.
  • NANOSECONDS - Time unit representing one thousandth of a microsecond.
  • SECONDS - Time unit representing one second.
typeenum
  • LEVEL - sets the first EMA value to the first value recorded.
  • DIFFERENCE - sets the first EMA value to zero assuming the change is initially zero.

Returns​

Parameterized constructor to use for calculating an EMA.

Examples​

In the following example, we use the time-table and update methods to create a table with random data to simulate real-time data. We then use ByEmaSimple to configure what we want for our EMA.

ema1sec is set to use BD_SKIP to skip null values and skip NaN values. The type is set to LEVEL and we are using TIME rather than TICKS.

ema_data then uses ema1sec to calculate the total EMA on a timescale of 10 seconds.

from deephaven.experimental.ema import ByEmaSimple
from deephaven import time_table
import random

def random_char():
    return random.choice(['A', 'B', 'C', 'D', 'E', 'F'])

ema1sec = ByEmaSimple('BD_SKIP','BD_SKIP','TIME',10,'SECONDS', type='LEVEL')

result = time_table('00:00:01').update(formulas=["Sym = (String)random_char()", "Numbers = (int)random.randint(0, 100)"])

ema_data = result.view(formulas=["EMA_data = ema1sec.update(Timestamp, Numbers)"]).tail(1)
ema_data_grouped = result.view(formulas=["Sym", "EMA_data = ema1sec.update(Timestamp, Numbers, Sym)"]).last_by(by=["Sym"]).tail(5)