Please note that the recommended version of Scilab is 2025.0.0. This page might be outdated.
See the recommended documentation of this function
nansum
Sum of values ignoring NAN's
Syntax
s = nansum(x,orient)
Arguments
- x
numerical vector or matrix.
- orient
nothing or '*'. 'r' or 1. 'c' or 2.
- s
Numerical scalar or vector containing the value of the adding operation.
Description
This function returns in s the sum of the values (ignoring the NAN's) of a numerical vector or matrix x.
For a vector or matrix x, s=nansum(x) (or s=nansum(x,'*')) returns in scalar s the sum of values of all entries (ignoring the NAN's) of a vector or matrix x.
s=nansum(x,'r')(or, equivalently, s=nansum(x,1)) returns in each entry of the row vector s of type 1xsize(x,'c') the sum of each column of x (ignoring the NANs).
s=nansum(x,'c')(or, equivalently, s=nansum(x,2)) returns in each entry of the column vector s of type size(x,'c')x1 the sum of each row of x (ignoring the NANs).
Examples
x=[0.2113249 %nan 0.6653811;0.7560439 0.3303271 0.6283918] m=nansum(x) m=nansum(x,1) m=nansum(x,2)
See also
Bibliography
Wonacott, T.H. and Wonacott, R.J.; Introductory Statistics, 5th edition, J.Wiley and Sons, 1990.
Report an issue | ||
<< nanstdev | Data with Missing Values | thrownan >> |