Unique not working - may be due to different datatypes in array

After smoothing out the data from one array I end up with one containing the values shown below. I want to get rid of the duplicate values 7 and 4.6 however unique will not get rid of them. Is this due to the fact that these values are different data types to each other? Any help would be appreciated.

Array

1.40000000000000

6.20000000000000

4.60000000000000

4.60000000000000

3.40000000000000

3.80000000000000

7.00000000000000

7

3.40000000000000

4

4.40000000000000

2.80000000000000

3.20000000000000

Answer 1

[Walter Roberson]

Those are not different datatypes.

When you see the same array displaying 7.00000000000000 and 7 both, then what it tells you is that the 7 value is exactly the floating point representation of 7, but that the 7.00000000000000 value is slightly different from 7 exactly. If you were to take Array(7) - Array(8) you would see a very small difference, probably roughly 2E-15 . The value that is exactly 7.0000000000000017763568394002504646778106689453125 displays as 7.00000000000000 in some cases because the default formats do not display all of the digits.

[James Morris]

Ah I see, that makes sense!

Knowing that, I just did:

array = round(array,2);

array = unique(array);

Is that a good work around?

[Walter Roberson]

It looks plausible for your situation.

[James Morris]

You're the man Walter, thank you.

[Stephen23]

You could also try uniquetol, which was designed to handle this situation.

https://www.mathworks.com/help/matlab/ref/uniquetol.html

[BN]

If any NaN values exist in rows, the unique function does not work (not eliminate parallel rows). I have this problem, so I converted NaN values to -99 using fillmissing, I use unique value again, and it's work.