Let’s say I have this:

```
Input1 Input2 Input3 Input4 Output1 Output2 Output3
1 1 1 1 4 20 11
1 12 7 14 18 11 8
4 2 18 7 0 6 16
```

The input values form a unique key (like a composite primary key in SQL). So each input column could have repeated values, but the whole set can’t repeat.

Then I have this for searching and displaying data:

```
Input1 Input2 Input3 Input4 Output1 Output2 Output3
A B C D E F G
```

`A`

to `D`

are the cells that I could key in. Based on these 4, `E`

to `G`

would find the correct row and display the correct output values.

I’ve tried using `vlookup`

as an array formula (CSE), but it is not giving me the correct answer. It is returning me the value on the last row that `Input1`

matches (regardless if other matched). Anyone knows how I can achieve this?

### Edit

This is the formula I tried for the cell in `E`

:

```
{=VLOOKUP(I2:L2, A2:G4, 5)}
```

where `I2:L2`

is the range with cells of values `A`

to `D`

, `A2:G4`

is the whole set of original data.

The formula evaluates to `18`

when the inputs are `1`

,`1`

,`1`

,`1`

, whereas this should match with the first row and evaluates to `4`

.

If A:D make unique combinations then you can use SUMIFS like this in M2,

```
=SUMIFS(E:E, $A:$A, $I2, $B:$B, $J2, $C:$C, $K2, $D:$D, $L2)
```

Drag right and down.

If A:D do not make unique combinations then you require a multi-column lookup, probably with INDEX/AGGREGATE; you also need to know which four-column match you want to retrieve.

Tags: excelexcel, search, sed