Question on why tilde cannot return the opposite statement result

Hi there, I want to compute the mean of working hours for those aged 30-50, here're code I wrote:

Q4_ans_test = mean( adultdata.hours_per_week( adultdata.age>30 & adultdata.age<=50))
Q4_ans_test2 = mean( adultdata.hours_per_week(adultdata.age>30 & ~adultdata.age>50),'omitnan')

The problem is the second statement returns "NaN". As I understand ,~ means the opposite of your following statement, and I expect to get the same answer as that in the first line of code. Anyone has idea on what's happening? Thank you so much!

Answer 1

[James Tursa]

Precedence of ~ is hgher than >, so it gets performed first. You need to use parentheses:

~(adultdata.age>50)

[Jing Gao]

It works. Thank you a lot for your help and explanation!

[Stephen23]

See also:

https://www.mathworks.com/help/matlab/matlab_prog/operator-precedence.html

Answer 2

[Walter Roberson]

~ binds more tightly than > does, so ~adultdata.age>50 is the same as (~adultdata.age)>50

Answer 3

[dpb]

Q4_ans_test2 = mean( adultdata.hours_per_week(adultdata.age>30 & ~(adultdata.age>50)));

Without the parentheses, the ~operator applies only to the variable adultdata.age and logical not applied to numeric converts anything nonzero to 1 so presuming nobody in the dataset is of age zero, the second includes all elements. There would still needs be a NaN in the mix somewhere, though, to return NaN as the result, but that's the reason why result is different; it is different set of data.

[Jing Gao]

I guess "l not applied to numeric converts anything nonzero to" 0 instead of 1? I've tried ~[1], it returns 0

[Walter Roberson]

For numeric values, ~X is (X == 0) . The exception is that ~nan is an error, whereas (nan == 0) is false

[Walter Roberson]

There would still needs be a NaN in the mix somewhere, though

No, with the ~ turning out false and the & condition being there, the logical condition was turning out false, so no elements were selected from the data, so mean() was being applied to an empty matrix. mean() of empty is NaN rather than empty.