Refining inefficient formula - for personal growth

[Whatamk]

I am rather proud of myself for the the following bit of formula, but with so many nested commands, I can't help but feel that I am doing this in a rather convoluted way. For the sake of growth, would you have approached this same question differently or more succinctly?

Drawing from this list of data...

	G	H	I	J	K	L	M	N	O	P
2	Axis I	Axis II	Axis III	Axis IV	Axis V	Blank	Dx ≈ FI	Ep. Dx	TP	Other
3	3	0	0	0	2	0	1	2	1	0
4	13	8	20	12	32	5	28	18	8	2
5	20	7	7	7	14	4	25	23	11	1
6	0	2	1	4	3	1	2	14	14	0

<tbody>
</tbody>

...I set about to create a list of the 3 highest categories (columns) for each row, resulting in...

	T	U	V
2	#1 error	#2 error	#3 error
3	Axis I	Axis V	Axis V
4	Axis V	Dx ≈ FI	Axis III
5	Dx ≈ FI	Ep. Dx	Axis I
6	Ep. Dx	Ep. Dx	Axis IV

<tbody>
</tbody>

As you can see, it's a little weird about equal numbers, but I can live with that. The formula I used was

=IF(SUM($G3:$P3)=0,"",INDIRECT(CONCATENATE(LEFT(CELL("address",INDEX($G3:$P3,MATCH(LARGE($G3:$P3,1),$G3:$P3,0))),2),"2")))

Any suggestions or thoughts?

 

[barry houdini]

To avoid repeats for duplicate numbers try this array formula in T3

=IFERROR(INDEX($G$2:$P$2,MATCH(1,($G3:$P3=LARGE($G3:$P3,COLUMNS($T3:T3)))*(COUNTIF($S3:S3,$G$2:$P$2)=0),0)),"")

confirmed with CTRL+SHIFT+ENTER and copied across and down.......I'm assuming column S is empty (or it can be populated but not with any value that might appear in G2:P2)

 

[Whatamk]

Works like a charm. Thank you, Barry! Definitely an improvement. I did make one modification: I put the =IF(SUM($G3:$P3)=0,"",(formula) back in rather than IFERROR so zero values returned blank rather than Axis I, Axis II, and Axis III.

As I hoped would happen, I'm a bit baffled by some of what you used (1: index/match combination is brilliant, and 2: I'm still trying to figure out why you used column S for the COUNTIF), so now I'm having to detour from work while I research it.