RollingCount
RollingCount calculates a window-based rolling count in an updateBy table operation. The rolling count can be calculated using forward and/or backward windows.
Syntax
RollingCount(revTicks, fwdTicks, pairs)
RollingCount(revTicks, pairs)
RollingCount(timestampCol, revTime, fwdTime, pairs)
RollingCount(timestampCol, revTime, pairs)
RollingCount(timestampCol, revDuration, pairs)
RollingCount(timestampCol, revDuration, fwdDuration, pairs)
Parameters
| Parameter | Type | Description |
|---|---|---|
| revTicks | long | The look-behind window size in ticks (rows). If positive, it defines the maximum number of rows before the current row that will be used. If negative, it defines the minimum number of rows after the current row that will be used. |
| fwdTicks | long | The look-forward window size in ticks (rows). If positive, it defines the maximum number of rows after the current row that will be used. If negative, it defines the minimum number of rows before the current row that will be used. |
| pairs | String... | The input/output column name pairs. |
| timestampCol | String | The name of the timestamp column. |
| revDuration | Duration | The look-behind window size in Duration. |
| fwdDuration | Duration | The look-forward window size in Duration. |
| revTime | long | The look-behind window size in nanoseconds. |
| fwdTime | long | The look-forward window size in nanoseconds. |
Returns
An UpdateByOperation to be used in an updateBy table operation.
Examples
The following example performs an updateBy on the source table using four row-based RollingCount operations. Each operation uses different revTicks and fwdTicks values to show how they affect the output. Note that the first operation only uses revTicks.
rng = new Random()
source = emptyTable(10).update("X = rng.nextInt(25)")
opRollCount = RollingCount(3, "RollingCount = X")
opPrior = RollingCount(3, -1, "WindowPrior = X")
opPosterior = RollingCount(-1, 3, "WindowPosterior = X")
opMiddle = RollingCount(1, 1, "WindowMiddle = X")
result = source.updateBy([opRollCount, opPrior, opPosterior, opMiddle])
- source
- result