Index Match with multiple results

Hi all,

I have two data sets; one is a client data set and the other is an episode data set.

Each client can have up to four programs that they can, legally, access at any one time.

What I need is to take the client code from the first data set and match it with the second data set that lists what program has been accessed by a specific client over the prior three months.


Both data sets update each month with tens of thousands of records in each data set.


What I want is to be able to produce a spreadsheet which says that client xxxxx has utilized these program(s) in the past three months. That may be one program, it could be two programs or up to four different programs.

If I could get the result to display in four separate columns (marked PRGCode1, PRGCode2 etc) that would be great. I tried that using the columns function but couldn't make it work.


I've attached a sample - you will see that my attempt at the formula is in column B but with mostly errors.


Any help in achieving the outcome would be greatly appreciated.

Standard question for situations like this: why aren't you using a database for this?

Given your sample workbook, try

B2:

Formula:

Please Login or Register  to view this content.

Copy B2, select B2:E16, and paste.

Hi hringrv,

that array formula has worked perfectly - and I don't know how.

In terms of using a database - the data comes from two different databases through an SQL query. I don't know enough about using a database such as Access to create an automated function that takes the data from these two queries and produces an output file every night for loading into the ERP.


Would still love to know how that formula works.

Originally Posted by Wizards

. . . Would still love to know how that formula works.

INDEX(($S$2:$S$20<>$A2)*1E+300+(ROW($S$2:$S$20)-CELL("row",$S$2:$S$20)+1),0) returns an array of very large numbers for values in S2:S20 which differ from the value of A2, or row indices in S2:S20 for values in S2:S20 equal to the value of A2. ROW($S$2:$S$20) is an array of row numbers, so {2;3;...;19;20}; CELL("row",$S$2:$S$20) is the row number of the 1st cell in S2:S20, so 2; ROW(...)-CELL(...) is then the array {0;1;...;17;18}, so adding 1 to it makes it the array {1;2;...;18;19}. COLUMNS($B2:B2) is the number of columns in its range argument, so this would be 1; however, as you fill or copy the formula to the right, it adjusts to COLUMNS($B2:C2), COLUMNS($B2:D2), etc, so returns 2, 3, etc. SMALL(p,q) returns the qth smallest value in p; those are row indices for values in S2:S20 equal to the value in A2 or very large numbers for values in S2:S20 differing from the value in A2. INDEX(x,y) returns the yth item in x for 1 <= y <= number of items in x; for very large y, INDEX(x,y) returns a #REF! error. IFERROR(INDEX(...),"") returns nonerror values and replaces error values with "" (zero length string, which appears as blank).