Calculated field in Pivot Table

[benb]

I have a dataset like the one below (not sure if formatting will come
out right in the posting). The example below is just a simplication of
my actual dataset.
Group Series Status
A A1 Open
A A2 Open
A A3 Unknown
A A9 Closed
A A10 Open
B B1 Closed
B B4 Closed
B B5 Closed
B B6 Closed
C C1 Unknown
D D1 Closed
D D2 Closed
E E1 Open
E E2 Open

What I want to do is count how many Groups fall into the status Open
bucket, how many into Closed, etc. Now because Group A has a Series in
each of the three possible Statuses, it falls into all three buckets.
Group E, on the other hand, only falls into one. If I pull Status into
a Pivot as a Column Field and then do a Count of Group, the values will
actually represent the number of Series in a given Status because the
value A appears in the dataset once for each corresponding Series in
that Group. Have I lost you yet? My hypothesis is that there is a
creative way to use calculated fields in a pivot to get it to show that
there are 2 groups showing Open, 2 showing Unknown, and 3 showing
Closed. I already know I could arrange the pivot in such a way that
=COUNTA( ) in the adjacent cells would give me the same thing, but I'm
looking for a way of doing it in the pivot for reason not worth wasting
any more space on. THANK IN ADVANCE FOR YOUR HELP!

[Debra Dalgleish]

You could add a column to the source data, and calculate if a record is
the first group/status item. For example,

=IF(SUMPRODUCT(--($A$2:$A2=A2),--($C$2:$C2=C2))=1,1,0)

Copy the formula to the last row of data.
Add the new field to the data area as a sum

benb wrote:
> I have a dataset like the one below (not sure if formatting will come
> out right in the posting). The example below is just a simplication of
> my actual dataset.
> Group Series Status
> A A1 Open
> A A2 Open
> A A3 Unknown
> A A9 Closed
> A A10 Open
> B B1 Closed
> B B4 Closed
> B B5 Closed
> B B6 Closed
> C C1 Unknown
> D D1 Closed
> D D2 Closed
> E E1 Open
> E E2 Open
>
> What I want to do is count how many Groups fall into the status Open
> bucket, how many into Closed, etc. Now because Group A has a Series in
> each of the three possible Statuses, it falls into all three buckets.
> Group E, on the other hand, only falls into one. If I pull Status into
> a Pivot as a Column Field and then do a Count of Group, the values will
> actually represent the number of Series in a given Status because the
> value A appears in the dataset once for each corresponding Series in
> that Group. Have I lost you yet? My hypothesis is that there is a
> creative way to use calculated fields in a pivot to get it to show that
> there are 2 groups showing Open, 2 showing Unknown, and 3 showing
> Closed. I already know I could arrange the pivot in such a way that
> =COUNTA( ) in the adjacent cells would give me the same thing, but I'm
> looking for a way of doing it in the pivot for reason not worth wasting
> any more space on. THANK IN ADVANCE FOR YOUR HELP!
>


--
Debra Dalgleish
Excel FAQ, Tips & Book List
http://www.contextures.com/tiptech.html

[benb]

Thanks Debra. That's a really good workaround that I would not have
caught and brings me close to an alternate solution for what I'm doing
without using a pivot at all.

Apart from practical concerns, my inquiry is one part stubborness, one
part learning for learning's sake. I am really unaccustomed to how
calculated fields work in pivots so I would like to learn more about
using them and I am convinced (perhaps wrongly so) that their
application can offer a solution here.

For now, I can do what I need to get done. If anyone has some insight
into how a calculated field might be used, though, I'm still very
curious to hear.

Thanks again.


Debra Dalgleish wrote:
> You could add a column to the source data, and calculate if a record is
> the first group/status item. For example,
>
> =IF(SUMPRODUCT(--($A$2:$A2=A2),--($C$2:$C2=C2))=1,1,0)
>
> Copy the formula to the last row of data.
> Add the new field to the data area as a sum
>
> benb wrote:
> > I have a dataset like the one below (not sure if formatting will come
> > out right in the posting). The example below is just a simplication of
> > my actual dataset.
> > Group Series Status
> > A A1 Open
> > A A2 Open
> > A A3 Unknown
> > A A9 Closed
> > A A10 Open
> > B B1 Closed
> > B B4 Closed
> > B B5 Closed
> > B B6 Closed
> > C C1 Unknown
> > D D1 Closed
> > D D2 Closed
> > E E1 Open
> > E E2 Open
> >
> > What I want to do is count how many Groups fall into the status Open
> > bucket, how many into Closed, etc. Now because Group A has a Series in
> > each of the three possible Statuses, it falls into all three buckets.
> > Group E, on the other hand, only falls into one. If I pull Status into
> > a Pivot as a Column Field and then do a Count of Group, the values will
> > actually represent the number of Series in a given Status because the
> > value A appears in the dataset once for each corresponding Series in
> > that Group. Have I lost you yet? My hypothesis is that there is a
> > creative way to use calculated fields in a pivot to get it to show that
> > there are 2 groups showing Open, 2 showing Unknown, and 3 showing
> > Closed. I already know I could arrange the pivot in such a way that
> > =COUNTA( ) in the adjacent cells would give me the same thing, but I'm
> > looking for a way of doing it in the pivot for reason not worth wasting
> > any more space on. THANK IN ADVANCE FOR YOUR HELP!
> >
>
>
> --
> Debra Dalgleish
> Excel FAQ, Tips & Book List
> http://www.contextures.com/tiptech.html