Cumulative Window Functions (cum-functions)

The cumulative window starts from a specified data/time point and aggregates data over consecutive data points or time intervals. DolphinDB provides the cumulative window functions (cum-functions) for cases involving cumulative windows.

Introduction

  • higher-order function accumulate:

    accumulate(func, X, [init])

    If init is specified, the result for the first window is init + X[0].

  • Syntax templates for the cum-functions:

    cumfunc(X)
    cumfunc(X, Y)

    Parameters:

    X (Y) is a scalar/vector/matrix/table/tuple with equal-length vectors.

Windowing Logic

The cum-function performs a cumulative calculation on each element, then returns a vector of the same length (or matrix of the same size) as the input.

The following example illustrates the calculation rules:

X = 1 1 2 3 4 5 NULL 8 9 1 3
cumsum(X)
// output
[1, 2, 4, 7, 11, 16, 16, 24, 33, 34, 37]