->count() VS select count(*) As count

They basically produce the same sql. Count(*) is a bit slower than count(active) or specifying a field, just like select a, b, c is faster than select *.

hmmm just for example, i know the column name, COUNT(id) count

and there is 5,000,000 rows in the collections

so is database return a count number VS use count(the array has 5,000,000 rows)

what the method ->count() used in Laravel , just use count(array()) ?

No, it's selecting the count in the query just as you are in your 2nd one. There's lots of shortcuts in eloquent. Like ->latest() produces "ORDER BY created_at DESC".

Install a tool like laravel debugbar, so you can see all of the actual queries being run. It's quite useful. https://github.com/barryvdh/laravel-debugbar

i have installed laravel-debugbar, tested on 350,000 rows data

CaseB is faster than CaseA

i think because in CaseA need to use php count(if many elements means need more time)

Then what are the queries both produce? You're 2nd one uses ->first(), which adds "LIMIT 1".

http://os33.atwebpages.com/laravel_debugbar.html

i captured the screen, the 1st query is CaseA, and 2nd is CaseB

oops, ->count() means use Count(*) in the query or use php count(result set) ?

It shows using count in the query, so it's not using php's count(result set). And yes, limit 1 will slightly make it faster as you're explicitly limiting the result set.

so means if i use CaseB is better than use CaseA ?

because i tested on a table has 350,000 rows data only

on another table has 3,200,000 rows, so i want to write a model, but donno which case is more better

It's up to you. There was only a 15ms difference in time, which is pretty trivial. So, it might take ~137ms, which is still pretty trivial. It's not even .25 seconds.

The difference seems negligible. I'd use the one that is easiest to work with. (Case A)

ok thankyou @Cronix and @BrandonSurowiec