Skip to main content
Version: Python

emmin_time

emmin_time creates a time-based EMMIN (exponential moving minimum) for an update_by table operation. The formula for the time-based EMMIN of a column XX is:

ai=e−dtiτa_i = e^{\frac{-dt_i}{\tau}}

min⁡0(X)=x0\min_0(X) = x_0

min⁡i(X)=min⁡(a∗min⁡i−1(X),  xi)\min_i(X) = \min(a*\min_{i-1}(X), \; x_i)

Where:

  • dtidt_i is the difference between time tit_i and ti−1t_{i-1} in nanoseconds.
  • τ\tau is decay_time in nanoseconds, an input parameter to the method.
  • min⁡i(X)\min_i(X) is the exponential moving minimum of XX at step ii.
  • xix_i is the current value.
  • ii denotes the time step, ranging from i=1i=1 to i=n−1i = n-1, where nn is the number of elements in XX.

Syntax​

emmin_time(
    ts_col: str,
    decay_time: Union[int, str],
    cols: list[str],
    op_control: OperationControl = None,
) -> UpdateByOperation

Parameters​

ParameterTypeDescription
ts_colstr

The name of the column in the source table containing timestamps.

decay_timeUnion[int, str]

The decay rate. This can be expressed as an integer in nanoseconds or a string duration, e.g., "PT00:00:00.001" or "PTnHnMnS", where H is hour, M is minute, and S is second.

colsUnion[str, list[str]]

The string names of columns to be operated on. These can include expressions to rename the output, e.g., "new_col = col". When this parameter is left empty, update_by will perform the operation on all applicable columns.

op_controlOperationControl

Defines how special cases should behave. When None, default OperationControl settings will be used. See OperationControl for information.

Returns​

An UpdateByOperation to be used in an update_by table operation.

Examples​

The following example performs an [update_by] on the source table using two emmin_time operations. Each uses a different decay_time value to demonstrate how it affects the output.

from deephaven.updateby import emmin_time
from deephaven import empty_table

source = empty_table(20).update(
    ["Timestamp = '2023-05-01T00:00:00 ET' + i * SECOND", "X = randomInt(0,25)"]
)

result = source.update_by(
    ops=[emmin_time(ts_col="Timestamp", decay_time="PT00:03:00.000", cols="EmMinX = X")]
)

result2 = source.update_by(
    ops=[emmin_time(ts_col="Timestamp", decay_time="PT00:01:00.000", cols="EmMinX = X")]
)