Date Formulas

[Guest]

I am trying to calculate the age of a person using the current "today's" date
and their birthdate without typing today's date into a cell. I do not need
the current date in a cell, due to having it in a footer. My thinking is
that it should look something like {=year(today)-year(birthdate)} I cannot
get this to work. I have their birthdates typed into a cell, but do not have
the current date.

 

[Biff]

See this:

http://cpearson.com/excel/datedif.htm

Biff

 

[Roger Govier]

Hi

If you only want the number of years, then you were almost there with
your formula
=YEAR(TODAY())-YEAR(A1)
where A1 is holding the birthdate.
You need to format the cell with the formula as General
Format>Cells>Number>General otherwise an age of say 44 would show as
13/02/1900

If you require the age split out into years, months days then follow
Biff's lead to Datedif, but do take note of some of its vagaries with
certain dates.

 

[Epinn]

Anyone who is interested in C. Pearson's link and wonders if it is faster to use TODAY( ) instead of NOW( ) in the age calculating formula, here is a thread for additional information.

http://groups.google.ca/group/micro...23022?lnk=st&q=&rnum=1&hl=en#4101c4fb16323022

Epinn

See this:

http://cpearson.com/excel/datedif.htm

Biff

 

[David Biddulph]

If you only want the number of years, then you were almost there with your
formula
=YEAR(TODAY())-YEAR(A1)
where A1 is holding the birthdate.
You need to format the cell with the formula as General
Format>Cells>Number>General otherwise an age of say 44 would show as
13/02/1900

That would give the age which will be achieved by the end of this year,
rather than the more usual usage which is the number of whole years already
achieved by today's date, which you could get by
=YEAR(TODAY()-(A1))-1900
(formatted as number with no decimal places).

 

[Roger Govier]

Hi David

You are absolutely correct. I had tested using my own birthday, which
has already occurred this year.
However, your formula will also give the wrong result when the month
today() is the same as the month(birthday)

=INT(YEARFRAC(birthday,today(),3)) (from the Analysis Toolpak, will
also give the wrong result because of leap years)
so maybe the best for whole year ages would be

=INT((Today()-Birthday)/365.25)

 

[David Biddulph]

Hi David

You are absolutely correct. I had tested using my own birthday, which has
already occurred this year.
However, your formula will also give the wrong result when the month
today() is the same as the month(birthday)

You've confused me there, Roger. I can't see why that should be the case
(as the calculation of the month doesn't come into my formula), and when I
test it with birthdays in October it shows the age changing for dates on or
before 12th October, as I would expect.

The only situation where I can foresee that there might be a difficulty is
around leap years, with the question of when someone born on 29th Feb
celebrates their birthday (without resorting to the Pirates of Penzance!).

 

[Roger Govier]

Hi David

Don't know what's going on here.
When I enter 11/10/1944 in A1
=YEAR(TODAY()-(A1))-1900
returns 61
30/09/1944 returns 62
but all October dates return 61

 

[David Biddulph]

I see what you mean, but are you sure that it's all dates in October, and
not just yesterday (11th Oct)? Doesn't 9/10/44 or 10/10/44 give 62?

I get the impression that we may just be a day adrift for some years, and my
guess is that it's tied up with leap years, but you're right that my simple
formula isn't 100% reliable. Thanks for pointing it out.
--
David Biddulph

Hi David

Don't know what's going on here.
When I enter 11/10/1944 in A1
=YEAR(TODAY()-(A1))-1900
returns 61
30/09/1944 returns 62
but all October dates return 61

 

[Roger Govier]

Hi David

Just tried it again and it returns 62 up to 10 Oct 1944 and 61
thereafter.
Not consistent with what I was seeing before.

I set up a table of dates from 30/09/44 through 11/11/44 down column A,
then filled through for 12 columns so I have from 30/09/44 to 11/11/56.
Copied the formula across as appropriate and then switch from 1 year to
one lower year seems to occur after 11/10 for years 44 and 45, 12/10 for
year 46 and 47, back to 11/10 years 48 and 49 and continues this pattern
of switching after every pair of years

It must be something to do with leap year and Rounding in Excel's
calculation.

I'm not going to give it any more thought!!!!

I have long since discovered that there is no foolproof method of
consistent calculation of date intervals other than the straight
subtraction of one date from another and returning the number of days.

"Sod" was obviously alive and well when our planetary system was formed
giving rise to a cycle around the Sun that is not an exact number of
days, let alone weeks and with a calendar devised which incorporates
such a disparate set of day ranges, how on earth can anyone expect to
get it right<vbg>

--
Regards

Roger Govier

I see what you mean, but are you sure that it's all dates in October,
and not just yesterday (11th Oct)? Doesn't 9/10/44 or 10/10/44 give
62?

I get the impression that we may just be a day adrift for some years,
and my guess is that it's tied up with leap years, but you're right
that my simple formula isn't 100% reliable. Thanks for pointing it
out.

 

[Epinn]

Do you want to know what function I prefer?

I still like DATEDIF even though I don't need the number of months and the number of days.

=DATEDIF(A1,TODAY(),"y")

I tested this formula on July which has passed, on December which is to happen and even on 10/12/2005. The formula passed all three tests and I got one year for October 12, 2005 (birthdate).

I hope MS won't take DATEDIF away in the future. Last night, I was wondering whether it is faster to use DATEDIF or to just do date2 - date1. I think I should give my brain a rest.

Epinn

Hi David

Just tried it again and it returns 62 up to 10 Oct 1944 and 61
thereafter.
Not consistent with what I was seeing before.

I set up a table of dates from 30/09/44 through 11/11/44 down column A,
then filled through for 12 columns so I have from 30/09/44 to 11/11/56.
Copied the formula across as appropriate and then switch from 1 year to
one lower year seems to occur after 11/10 for years 44 and 45, 12/10 for
year 46 and 47, back to 11/10 years 48 and 49 and continues this pattern
of switching after every pair of years

It must be something to do with leap year and Rounding in Excel's
calculation.

I'm not going to give it any more thought!!!!

I have long since discovered that there is no foolproof method of
consistent calculation of date intervals other than the straight
subtraction of one date from another and returning the number of days.

"Sod" was obviously alive and well when our planetary system was formed
giving rise to a cycle around the Sun that is not an exact number of
days, let alone weeks and with a calendar devised which incorporates
such a disparate set of day ranges, how on earth can anyone expect to
get it right<vbg>

--
Regards

Roger Govier

I see what you mean, but are you sure that it's all dates in October,
and not just yesterday (11th Oct)? Doesn't 9/10/44 or 10/10/44 give
62?

I get the impression that we may just be a day adrift for some years,
and my guess is that it's tied up with leap years, but you're right
that my simple formula isn't 100% reliable. Thanks for pointing it
out.

 

[Epinn]

Further test with my formula ......

10/15/2005 yields 0 whereas 10/15/2004 yields 1. I think this is acceptable.

I won't go into leap year here as I started a thread on DATEDIF and leap year yesterday.

Thank you for reading.

Epinn

Do you want to know what function I prefer?

I still like DATEDIF even though I don't need the number of months and the number of days.

=DATEDIF(A1,TODAY(),"y")

I tested this formula on July which has passed, on December which is to happen and even on 10/12/2005. The formula passed all three tests and I got one year for October 12, 2005 (birthdate).

I hope MS won't take DATEDIF away in the future. Last night, I was wondering whether it is faster to use DATEDIF or to just do date2 - date1. I think I should give my brain a rest.

Epinn

Hi David

Just tried it again and it returns 62 up to 10 Oct 1944 and 61
thereafter.
Not consistent with what I was seeing before.

I set up a table of dates from 30/09/44 through 11/11/44 down column A,
then filled through for 12 columns so I have from 30/09/44 to 11/11/56.
Copied the formula across as appropriate and then switch from 1 year to
one lower year seems to occur after 11/10 for years 44 and 45, 12/10 for
year 46 and 47, back to 11/10 years 48 and 49 and continues this pattern
of switching after every pair of years

It must be something to do with leap year and Rounding in Excel's
calculation.

I'm not going to give it any more thought!!!!

I have long since discovered that there is no foolproof method of
consistent calculation of date intervals other than the straight
subtraction of one date from another and returning the number of days.

"Sod" was obviously alive and well when our planetary system was formed
giving rise to a cycle around the Sun that is not an exact number of
days, let alone weeks and with a calendar devised which incorporates
such a disparate set of day ranges, how on earth can anyone expect to
get it right<vbg>

--
Regards

Roger Govier

I see what you mean, but are you sure that it's all dates in October,
and not just yesterday (11th Oct)? Doesn't 9/10/44 or 10/10/44 give
62?

I get the impression that we may just be a day adrift for some years,
and my guess is that it's tied up with leap years, but you're right
that my simple formula isn't 100% reliable. Thanks for pointing it
out.