#NUM! error in a simple math equation

I am not sure why Excel's programmers did it, but, according the the help file: https://support.office.com/en-us/art...2-633196f6fdcc If x<0, Weibull returns the #NUM error. The error you are getting from the Weibull function is built into the function by Excel's programmers.

You may need to know exactly what the math means here (and I might suggest that you check your math carefully here). I would note that (-1E5/693)^4.25 is not the same as -1*(1E5/693)^4.25. The first expression is trying to take the 4th root of a negative number (which is not a real number, hence the #NUM error) and the second expression is taking the 4th root of a positive number (which is real), and then multiplying by -1. I expect that the same thing happens in the Weibull CDF where x cannot be negative (and get a real result). If you are expecting a real result, you will need to think through the parameters and make sure you are feeding the correct parameters into the Weibull function (whether Excel's built in function or your own version).

If you are expecting a complex result, you can try using the IMPOWER() function, which can raise a negative number to a fractional power, though it still gets a little tricky as it will only give one of the many possible results.

That should explain the #Num error, but I'm not sure it answers the question fully. What are you expecting?

However I have #NUM! error if i write "=-(1E5/693)^4.25".

This is one of those places where Excel is not consistent with typical math notation and other programming languages. In Excel's order of operations, negation (the leading minus sign) comes before exponentiation (https://support.office.com/en-us/art...8-7af9e0e2878a ). You will need to either multiply by -1, or make sure the exponentiation is inside of parentheses in order to make sure the negation comes after the exponentiation.

One thing I would suggest -- check your math and function definition carefully. Looking at the CDF given in Excel's help file for the Weibull function (1-exp(-1*(x/b)^a)), your Weibull function is using 100 for x, 693 for a, and 4.25 for b -- 1-exp(-1*(100/4.25)^693). Just about anything raised to the 693 power will exceed the limits of double precision and generate a #NUM error. Did you intend 4.25 to be a and 693 to be b =WEIBULL(100;4,25;693;true)?