Ron,
I agree and gave a link to a thread that pointed out those anomolies but the
real point for me is why anyone wants to include months in an age calculation
in the first place because the varying length makes it a meaningless number.
I may be wrong but my guess is that Microsoft never documented datedif in
Excel because it's quite the silliest function they ever produced.
Mike
> >Hello
> >Let's Suppose i gave some date 01-01-2007 in cell A1 and 01-05-2010 in cell b1
[quoted text clipped - 29 lines]
> time, especially if your starting date is not at the end of the month.
> --ron
Ron Rosenfeld - 17 May 2008 12:44 GMT
>Ron,
>
[quoted text clipped - 6 lines]
>
>Mike
Well, it can certainly give inconsistent results.
--ron
Fred Smith - 17 May 2008 14:34 GMT
You may be right Mike, but it still doesn't absolve Microsoft. As you can
see by the post, people want to express age differences in months. You see
it on TV all the time. By producing the function, Microsoft was simply
responding to customer demand. If they are going to produce it, they should
document it.
Regards,
Fred.
> Ron,
>
[quoted text clipped - 47 lines]
>> time, especially if your starting date is not at the end of the month.
>> --ron
Hello Ron,
For the 3 examples you gave Mike's formula gives me 29, 28 and 1 day
respectively which appear to me to be correct, although I think you get odd
results if start date is 31st January and end date 1st March.....in which
case try this formula
=YEAR(B1)-YEAR(A1)-(TEXT(B1,"mmdd")< TEXT(A1,"mmdd"))&" years
"&MOD(MONTH(B1)-MONTH(A1)-(DAY(B1)< DAY(A1)),12)&" months
"&B1-MIN(DATE(YEAR(B1),MONTH(B1)-(DAY(B1)< DAY(A1))+{1,0},DAY(A1)*{0,1}))&"
days"
> >Hello
> >Let's Suppose i gave some date 01-01-2007 in cell A1 and 01-05-2010 in cell b1
[quoted text clipped - 29 lines]
> time, especially if your starting date is not at the end of the month.
> --ron
Ron Rosenfeld - 18 May 2008 00:36 GMT
>Hello Ron,
>
>For the 3 examples you gave Mike's formula gives me 29, 28 and 1 day
>respectively which appear to me to be correct, although I think you get odd
>results if start date is 31st January and end date 1st March.....in which
>case try this formula
Those are odd results, since I get something quite different. Something
strange is going on.
Obviously you didn't copy and paste your results, since Mike's formula gives a
string.
Here is what I get -- copied and pasted:
A B C
31-Jan-2008 29-Feb-2008 0 years 0 months 27 days
31-Jan-2007 28-Feb-2007 0 years 0 months 25 days
29-Feb-2008 1-Mar-2008 0 years 0 months 3 days
With Mike's formula, also copied and pasted (but dragged down from C1):
=DATEDIF(A1,B1,"y")&" years "&DATEDIF(A1,B1,"ym")&
" months "&DATEDIF(A1,B1,"md")&" days"
It'll be very interesting if this function gives different results in different
versions of Excel. I'm using Excel 2007
>I think you get odd results if start date is 31st January and end date 1st March
Here's what I get using Mike's formula:
31-Jan-2008 1-Mar-2008 0 years 1 months 1 days
That seems like a perfectly reasonable answer. And it is also the same as the
answer I get using your formula.
--ron