Hi again,

Got it working if bringing in the ISNUMBER check twice. I don't know if this
is the best way, but it works. Thanks for the hint.

{=SUMPRODUCT(--(MOD(COLUMN(FM3:FR3);2)=0);IF(ISNUMBER(FM3:FR3);FM 3:FR3;""))/SUMPRODUCT(--(MOD(COLUMN(FM3:FR3);2)=0);(IF(ISNUMBER(FM3:FR3);1 ;0)))}

--
Regards Frode


"Frode" wrote:

Thanks Stefi, but this returned #N/A for the 2nd row where one of the data
fields was #N/A. It worked for the rows where all data fields were valid.

--
Regards Frode


"Stefi" wrote:

Sorry, this is the correct version for AVG:

=SUMPRODUCT(--(MOD(COLUMN(A2:F2),2)=0),A2:F2)/SUMPRODUCT(--(MOD(COLUMN(A2:F2),2)=0),--(ISNUMBER(A2:F2)))


--
Regards!
Stefi



€žStefi€� ezt Ã*rta:

One way:
=SUMPRODUCT(--(MOD(COLUMN(A2:G2),2)=0),A2:G2)/SUMPRODUCT(--(MOD(COLUMN(A2:G2),2)=0),--(ISNUMBER(A2:G2)))
Fill it down!
--
Regards!
Stefi



€žFrode€� ezt Ã*rta:

Hi,

I have several data tables, where every second column is an index and the
other columns uses the index to fetch the corresponding data. If any data is
invalid, #N/A is returned, so that the data can be shown in graphs (used ""
instead of #N/A in the beginning, but this made the graphs very noisy,
dropping to zero).
For each column of data the MIN, MAX and AVG is given by an array formula,
e.g.
{=AVERAGE(IF(ISNUMBER(B11:B754);B11:B754;""))}.
So far so good.
The problem is to give the average of each row in the tables, the RowAVG
below, since the data are in non-consequtive columns.
I would be greatful to receive a solution to this problem, without having to
duplicate the Data columns into a similar table but without the Index
columns.

Index1 Data1 Index2 Data2 Index3 Data3 RowAVG
1 10 1 20 1 30
20
2 15 2 #N/A 2 35
25
3 35 3 20 3 65
40

ColAVG 20 20 43.3
ColMIN 10 20 30