Scilab Website | Contribute with GitLab | Mailing list archives | ATOMS toolboxes
Scilab Online Help
5.4.1 - English

Change language to:
Français - 日本語 - Português - Русский

Please note that the recommended version of Scilab is 2025.0.0. This page might be outdated.
See the recommended documentation of this function

Scilab help >> Elementary Functions > Matrix operations > cumsum

cumsum

cumulative sum of array elements

Calling Sequence

y=cumsum(x)
y=cumsum(x,orientation)
y=cumsum(x,outtype)
y=cumsum(x,orientation,outtype)

Arguments

x

an array of reals, complex, booleans, polynomials or rational fractions.

orientation

This argument can be

  • either a string with possible values "*", "r", "c" or "m"

  • or a number with positive integer value

outtype

a string with possible values "native" or "double".

y

scalar or array

Description

For an array x, y=cumsum(x) returns in the scalar y the cumulative sum of all the elements of x.

y=cumsum(x,orientation) returns in y the cumulative sum of x along the dimension given by orientation:

  • if orientation is equal to 1 or "r" then:

    or

  • if orientation is equal to 2 or "c" then:

    or

  • if orientation is equal to n then:

  • y=cumsum(x,"*") is equivalent to y=cumsum(x)

  • y=cumsum(x,"m") is equivalent to y=cumsum(x,orientation) where orientation is the index of the first dimension of x that is greater than 1. This option is used for Matlab compatibility.

The outtype argument rules the way the summation is done:

  • For arrays of floats, of polynomials, of rational fractions, the evaluation is always done using floating points computations. The "double" or "native" options are equivalent.

  • For arrays of integers,

    if outtype="native" the evaluation is done using integer computations (modulo 2^b, where b is the number of bits used),

    if outtype="double" the evaluation is done using floating point computations.

    The default value is outtype="native".

  • For arrays of booleans,

    if outtype="native" the evaluation is done using boolean computations ( + is replaced by |),

    if outtype="double" the evaluation is done using floating point computations (%t values are replaced by 1 and %f values by 0).

    The default value is outtype="double".

Remark

This function applies, with identical rules to sparse matrices and hypermatrices.

Examples

A=[1,2;3,4];
cumsum(A)
cumsum(A,1)

I=uint8([2 95 103;254 9 0])
cumsum(I) //native evaluation
cumsum(I,"double")
cumsum(I,2,"double")

s=poly(0,"s");
P=[s,%i+s;s^2,1];
cumsum(P),
cumsum(P,2)

B=[%t %t %f %f];
cumsum(B) //evaluation in float
cumsum(B,"native") //similar to or(B)

See Also

  • sum — sum of array elements
  • cumprod — cumulative product of array elements
Report an issue
<< cumprod Matrix operations kron >>

Copyright (c) 2022-2024 (Dassault Systèmes)
Copyright (c) 2017-2022 (ESI Group)
Copyright (c) 2011-2017 (Scilab Enterprises)
Copyright (c) 1989-2012 (INRIA)
Copyright (c) 1989-2007 (ENPC)
with contributors
Last updated:
Tue Apr 02 17:36:21 CEST 2013