Laravel Eloquent grab duplicates (2 or more results) - php

Question

So I have the following query, which is working:
Expense::select('amount', 'date', DB::raw('COUNT(*) as `count`'))
->groupBy('amount', 'date')
->havingRaw('COUNT(*) > 1')
->get();
This is to grab all expenses that are potential duplicates (same amount and same date). However, I want to grab everything, not only the amount and date. The following is not working:
return Expense::select('name', 'amount', 'date', DB::raw('COUNT(*) as `count`'))
->groupBy('amount', 'date')
->havingRaw('COUNT(*) > 1')
->get();
This will give me this error:
SQLSTATE[42000]: Syntax error or access violation: 1055 Expression #1 of the SELECT list is not in the GROUP BY clause and contains a nonaggregated column
I have many more fields for an Expense that I want to list in my view, most importantly of course the name, but also the slug so I can link to each potential duplicate, etc.
Also, the second "issue", but not the most important one, is that with the above query, all expenses are returned in one collection. They are not grouped by date for example.
The most desirable would be to have the results somewhat like this. All potential duplicates should be grouped by the date so I can do something like this in my view:
#foreach($duplicates as $date => $expenses)
#foreach($expenses as $expense)
{{ $date }}
// List all expenses for that date.. etc.
Question 1: How would I make the above query work with all database fields, not only the date and amount?
Question 2: How would I be able to grab the potential duplicates and group them by date so I can loop over the dates and the expenses like in the above example?

Answer 1

Always handle column properly when doing grouping
e.i. you can specify all the fields you want to select in groupBy
// database query
$expenses = Expense::select(['amount', 'date', 'name', 'slug'])
->selectRaw('count(*) AS count')
->havingRaw('count > 1')
->groupBy(['amount', 'date', 'name', 'slug'])
->get();
// Another date grouping done in collection
return $expenses->groupBy('date');
or something like combing columns that belongs to your desired grouping
return Expense::select('date')
->selectRaw('group_concat(name) as names') // comma separated names belongs to the group
->selectRaw('group_concat(amount) as amounts') // comma separated amounts belongs to the group
->selectRaw('group_concat(slug) as slugs')// comma separated slugs belongs to the group
->selectRaw('count(*) AS count')
->havingRaw('count > 1')
->groupBy('date')
->get();
EDIT
If you only want to group them by date and still have the actual rows of duplicate, you can just add another select statement to count for date duplicate without grouping them by date in your query, and do the date grouping in collection.
e.i.
$expenses = Expense::select( ['amount', 'date', 'name', 'slug'] )
->selectRaw('(SELECT COUNT(date) FROM expenses t1 WHERE t1.date = expenses.date ) as duplicates')
->havingRaw('duplicates > 1')
->get();
return $expenses->groupBy('date');
Then the output should have another column duplicates which has the number of duplicate each date has and still has all the rows for duplicate dates