Validate for unique composite key in an excel using macro - vba

Question

I am creating an excel macro. As part of that I need to validate unique composite key in an Excel . ie say Column1 + Column2 + Column3 shouldn't be repeating.
If so the row should be highlighted in red color.
What is the best way to do it?
Thanks in Advance

Answer 1

There are several ways to accomplish this: Two depend on sorting your data while the third does not. I'll supply them in different answers so readers can indicate which they prefer.
Identify duplicates using a Dictionary in VBA
Pro: Fast, does not require sorting
Con: Requires code, does not automatically update
In this case I treat the problem of actually identifying the duplicate rows separately from the trivial step of highlighting them. This function returns a Dictionary where the keys are the compound keys that have more than one row and the values are Collections containing the row numbers of all the rows which matched the key. It's the equivalent of a Dictionary<string,List<int>> in .NET. Conceptually it looks like this:
"some..key..1" : [1, 42, 401]
"some..key..2" : [134, 135]
The keys are the concatenated contents of each of the key columns, delimited by a null character. I use the unprintable null character so that the key set ("A", "Dog", "2") does not become equal to ("AD", "o", "g2").
As written the key comparison is case-sensitive. If you desire a case-insensitive match, set the CompareMode property of dctValues and dctDuplicates to TextCompare.
Note: You'll need to add a reference to Microsoft Scripting Runtime
Public Function FindDuplicates(ByVal DataRange As Range, ParamArray KeyColumns()) As Dictionary
Dim ws As Worksheet
Dim vKeyRange, rngCol As Range
Dim dctKeys As New Dictionary
Dim colKeys
Dim keyParts() As String
Dim strKey As String
Dim dctValues As New Dictionary
Dim dctDuplicates As New Dictionary
Dim i As Long, ub As Long
Dim lngFirstRow As Long, lngLastRow As Long, lngRow As Long
Set ws = DataRange.Worksheet
' Identify unique key column numbers
For Each vKeyRange In KeyColumns
For Each rngCol In vKeyRange.Columns
dctKeys(rngCol.Column) = True
Next
Next
colKeys = dctKeys.Keys
ub = UBound(colKeys)
ReDim keyParts(ub)
' Find first and last row of data range
lngFirstRow = DataRange.Cells(1, 1).Row
lngLastRow = DataRange.Cells(DataRange.Rows.Count, 1).Row
' Loop through rows
For lngRow = lngFirstRow To lngLastRow
' Get the parts for the key
For i = 0 To ub
keyParts(i) = ws.Cells(lngRow, colKeys(i)).Value
Next
' Concatenate the parts with an unprintable character as
' the delimiter, so that "A" + "Dog" != "AD" + "og"
strKey = Join(keyParts, Chr(0))
' If the key hasn't been found yet, create a new collection
If Not dctValues.Exists(strKey) Then
dctValues.Add strKey, New Collection
End If
' Push the row number to the list of rows with this key
dctValues(strKey).Add lngRow
' If this is the second row with this key, add the
' list to the dictionary of keys with multiple rows
If dctValues(strKey).Count = 2 Then
dctDuplicates.Add strKey, dctValues(strKey)
End If
Next
Set FindDuplicates = dctDuplicates
End Function
Usage: Find all duplicate rows in A2:I5000, using columns A, B, and E as key columns
Dim ws As Worksheet, dctDups As Dictionary, vKey, vRow
Set ws = ThisWorkbook.Worksheets(1)
Set dctDups = FindDuplicates(ws.Range("A2:I5000"), ws.Range("A:B"), ws.Range("E:E"))
For Each vKey In dctDups
For Each vRow In dctDups(vKey)
ws.Range("A" & vRow & ":I" & vRow).Interior.Color = vbRed
Next
Next

Answer 2

There are several ways to accomplish this: Two depend on sorting your data while the third does not. I'll supply them in different answers so readers can indicate which they prefer.
Sort and apply conditional formatting
Pro: Dynamic (adjusts to changes in data), does not require any code
Con: Requires sorting, can become messy
Manually sort by the key columns
Create a conditional formatting rule and apply it to all rows of data.
Highlight all the data, but starting with the first row of data
Select Conditional Formatting -> New Rule
Set the format to a red fill
Select "Use a formula to determine which cells to format"
Here's the formula you need, assuming your selection starts on row 2 (there's a header in row 1), and your key columns are A, B, and C. Note carefully where the $ signs appear and where they do not:
=OR((CONCATENATE($A2,$B2,$C2)=CONCATENATE($A1,$B1,$C1)),
(CONCATENATE($A2,$B2,$C2)=CONCATENATE($A3,$B3,$C3)))
This will highlight both rows that have duplicate keys, or all rows if there are more than two.