I have two csv files with following columns
Csv1
Name| address| phone
John| Blah | Blah
Dave| Blah | Blah
Rick| Blah | Blah
Csv2
Name |address| phone
John| Blah| Blah
Dave| Blah| Blah
I want unique rows from first csv
What i want
Name | address| phone
Rick | Blah | Blah
My code
$csv1 = import-csv <file> $csv2 = import-csv <file>
$output = $csv1 | where-object {$_.name -notin $csv2.Name -and $_.address -notin $csv2.address -and $_.phone -notin $csv2.phone}
Your statement
| where-object {$_.name -notin $csv2.Name -and $_.address -notin $csv2.address -and $_.phone -notin $csv2.phone}
The logic when evaluating Rick's row
$Csv2 ? (Answer:No)$Csv2 ? (Answer: Yes)$Csv2 ? (Answer: Yes)You are evaluating every single row against all of the rows of $csv2 so when you check Rick's address ($_.address -notin $csv2.address), it is actually in $Csv2 for John and Dave. Based on how you formulated your condition, that does eliminate Rick from the unique rows.
You can use Compare-Object to compare both csv files and find items that aren't a match. Here is a demonstration comparing both csv using the 3 properties as criterias.
Compare-Object -ReferenceObject $csv1 -DifferenceObject $csv2 -Property Name,Address,phone | Where-Object SideIndicator -eq '<='
Complete self-contained example
$Csv1 = @'
Name|address|phone
John|Blah|Blah
Dave|Blah|Blah
Rick|Blah|Blah
'@ | ConvertFrom-Csv -Delimiter '|'
$csv2 = @'
Name|address|phone
John|Blah|Blah
Dave|Blah|Blah
'@ | ConvertFrom-Csv -Delimiter '|'
$Results = Compare-Object -ReferenceObject $csv1 -DifferenceObject $csv2 -Property Name,Address,phone | Where-Object SideIndicator -eq '<='
$Results | Select Name,Address,Phone
Output
Name Address phone
---- ------- -----
Rick Blah Blah
