Question on number accuracy and indexing

hi all, here is my problem. In the following code, I first give sigma an array from 0 to 1 with a step of 0.01. However, if I try to retrieve some value from it by indexing in the following code, I get '5.600000000000001e-01' for the first equation and an 'Empty matrix: 1-by-0' for the second. I know this must be related with the floating-point accuracy, but how can I avoid the empty matrix?

            sigma=[0:0.01:1];
            sigma(sigma==0.56) 
            sigma(sigma==0.57)

Answer 1

[Matt Fig]

This is the old Floating Point FAQ. Pay particular attention to the section where using a tolerance is discussed.

[Guang]

thanks for the reference.