MATLAB: NaN for zero divide by zero.

nan

Will illustrate with an example, I.e:

a = (0:0.1:1); b = (1-a)*2./(1-a);

Matlab will output the end value in b as NaN, when in reality b should converge to 2 for a = 1.

How can I avoid this?

Best Answer

There are two things to remember when looking at this problem.

1. A function having a limit as the independent variable approaches a point does not imply that the function value at that point is defined. You plugged in for the function value at that point. The fact that you get NAN says nothing about the limit of the function as the independent variable approaches that point.

2. NAN is the correct value when evaluating the function at that point. This is a standard double precision definition.

Related Solutions

MATLAB: Ignore NaN values when sorting

validmask = ~isnan(A);
[~, nonnanrank] = sort(A(validmask));
rank = zeros(size(A));
rank(validmask) = nonnanrank;

However I recommand that you use a different variable name rather than rank to avoid interfering with using the MATLAB function rank() . rank() is used less frequently than sum() or length() are, so this is perhaps less of an issue than using those variables would be.

MATLAB: How does particle swarm deal with objective function evaluations to NaN

The default value is to let the NaN "pollute" whatever calculation it is involved in. For example,

(nan+2+5)/3

is nan. Any arithmetic operation involving NaN and floating point numbers will result in NaN (if the integer classes are involved the answer is weirder.) Logically any comparison involving nan is false except for ~= which is defined as true even if both values are nan. This can have unexpected effects in the calculations that depend on exactly how the calculation is code.

A min() involving nan will only result in nan if all of the values are nan, so a particle with nan will not be detected as the best value. On the other hand its updated position could become nan, potentially the positions of other particles could become nan if some kind of consensus were being used, and so on. Depending on the coding, it might get replaced by a better result and have no further effect, or it might just be unproductive, or it might poison the calculations and mess up other particles.

Best Answer

Related Solutions

MATLAB: Ignore NaN values when sorting

MATLAB: How does particle swarm deal with objective function evaluations to NaN

Related Question