Why does Matlab incorrectly return 0 when evaluating an array, but correctly return NaN when evaluating a scalar?

Question

Why does the expression for h(theta) below sometimes return 0 and sometimes return NaN? h(theta) contains a division by zero for theta = 0 and should always return NaN. If I just ask for h(0), it all works fine.

However, when it evaluates a zero contained in a multi-element array, it returns h = 0 when it should return NaN. But if specifically evaluating only the element that is zero, then it returns NaN as it should.

>> theta = [0 1]
theta =
     0     1

The first element should be NaN:

>> h = tan(theta)/(1+(tan(theta)/tan(2.*theta)))
h =
     0    5.4220

When evaluating the zero element specifically it works correctly:

>> h = tan(theta(1))/(1+(tan(theta(1))/tan(2.*theta(1))))
h =
     NaN 

>> h = tan(theta(2))/(1+(tan(theta(2))/tan(2.*theta(2))))
h =
     5.4220

You can put >> to the input lines for formatting and better readability. — percusse
– percusse, Commented Jul 27, 2015 at 11:38

Matt · Accepted Answer · 2015-07-27 11:36:07Z

5

You need to do an element-wise division with ./ instead of / to get the desired result. This is called the right array division.

>> h = tan(theta)./(1+(tan(theta)./tan(2.*theta)))
h =
       NaN    5.4220

answered Jul 27, 2015 at 11:36

Matt

14.1k2 gold badges39 silver badges58 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Carl Witthoft Over a year ago

File this under "one more thing MATLAB does contrary to how every other language does it, and contrary to normal expectations"

Collectives™ on Stack Overflow

Why does Matlab incorrectly return 0 when evaluating an array, but correctly return NaN when evaluating a scalar?

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related