I'm new to MySQL and databases and I've seen in many places that it is not considered a good programming practice to use subqueries in the FROM field of SELECT in MySQL, like this:
select userid, avg(num_pages)
from (select userid , pageid , regid , count(regid) as num_pages
from reg_pag
where ativa = 1
group by userid, pageid) as from_query
group by userid;
which calculates the average number of registers per page that the users have.
The reg_page table looks like this:
Questions:
How to change the query so that it doesn't use the subquery in the FROM field?
And is there a general way to "flatten" queries like this?
The average number of registers per page per user can also be calculated as number of registers per user divided by number of pages per user. Use count distinct to count only distinct psgeids per user:
select userid, count(regid) / count(distinct pageid) as avg_regs
from reg_pag
where ativa=1
group by userid;
There is no general way of flattening such queries. It may not even be possible to flatten some of them, otherwise there would be little point in having this feature in the first place. Do not get scared of using subqueries in the from clause, in some occasions they may be even more effective, than a flattened query. But this is vendor and even version specific.
One way is to use count(distinct):
select userid, count(*) / count(distinct pageid)
from reg_pag
where ativa = 1
group by userid;
Related
I am working on a query that needs to output 'total engagements' by users in columns like 1 -eng column will display users who have one engagements, second column 2-eng which will display users who have done 2 engagements. Likewise 3eng, and so on. Note that the display should be like this. I have a engagement table which has userID. So I get distinct users like this
select count(distinct userID) from engagements
and I get engagements as
select count(*) from engagements
Engagements here refers to users who have either liked,replied,or shared the content
Please help. Thanks! I have used CASE and IF but unable to display in the below form
1eng 2eng 3eng
100 200 100
Consider returning the results in rows and pivoting them afterwards in your application.
To return the desired results in rows, you could use the following query:
SELECT
engagementCount,
COUNT(*) AS userCount
FROM (
SELECT
userID,
COUNT(*) AS engagementCount
FROM engagements
GROUP BY userID
) AS s
GROUP BY engagementCount
;
Basically, you first group the engagements rows by userID and get the row counts per userID. Afterwards, you use the counts as the grouping criterion and count how many users were found with that count.
If you insist on returning the columnar view in SQL, you'll need to resort to dynamic SQL because of the indefinite number of columns in the final result set. You'd probably need to store the results of the inner SELECT temporarily, scan it to build the list of count expressions for every engagementCount value and ultimately construct a query of this kind:
SELECT
COUNT(engagementCount = 1 OR NULL) AS `1eng`,
COUNT(engagementCount = 2 OR NULL) AS `2eng`,
COUNT(engagementCount = 3 OR NULL) AS `3eng`,
...
FROM temporary_storage
;
Or SUM(engagementCount = value) instead COUNT(engagementCount = value OR NULL). (For me, the latter expresses the intention more explicitly, hence why I've suggested it first, but, in case you happen to prefer the SUM technique, there should be no discernible difference in performance between the two. The OR NULL trick is explained here.)
We have a vehicles table that has a VehicleName column. I'm looking into a bug where a user has multiple vehicles of the same VehicleName. I have this query to help return which VehicleNames have been used multiple times:
SELECT VehicleName,
(
SELECT count(VehicleName)
FROM Vehicles as V2
WHERE V1.VehicleName = V2.VehicleName
)
FROM Vehicles as V1;
For one thing, it's slow. That's not too bad because this isn't going into production; it's just to aid in a bug fix. Second, this will return every VehicleName, even the ones that have a count of one, those are VehicleNames I'm not interested in for this application. I can't remember how to name the subquery, so I can't add a where to limit it.
I'm interested not just in how to name the subquery, but also are their faster solutions to this?
Are you trying to do this in some ugly way ?
SELECT VehicleName, COUNT(*) as TOTAL
FROM Vehicles
GROUP BY VehicleName
HAVING TOTAL > 1
ORDER BY TOTAL DESC;
This sounds quite simple but I just can't figure it out.
I have a table orders (id, username, telephone_number).
I want to get number of orders from one user by comparing the last 8 numbers in telephone_number.
I tried using SUBSTR(telephone_number, -8), I've searched and experimented a lot, but still I can't get it to work.
Any suggestions?
Untested:
SELECT
COUNT(*) AS cnt,
*
FROM
Orders
GROUP BY
SUBSTR(telephone_number, -8)
ORDER BY
cnt DESC
The idea:
Select COUNT(*) (i.e., number of rows in each GROUPing) and all fields from Orders (*)
GROUP by the last eight digits of telephone_number1
Optionally, ORDER by number of rows in GROUPing descending.
1) If you plan to do this type of query often, some kind of index on the last part of the phone number could be desirable. How this could be best implemented depends on the concrete values stored in the field.
//Memory intensive.
SELECT COUNT(*) FROM `orders` WHERE REGEXP `telephone_number` = '(.*?)12345678'
OR
//The same, but better and quicker.
SELECT COUNT(*) FROM `orders` WHERE `telephone_number` LIKE '%12345678'
You can use the below query to get last 8 characters from a column values.
select right(rtrim(First_Name),8) FROM [ated].[dbo].[Employee]
select id, first_name, count(*) from users;
The users table contains 10 entries, but the above select query shows only a single row. Is it wrong to mix count(*) as part of the above query?
COUNT is a function that aggregates. You can't mix it into your normal query.
If you want to receive the ten entries just do a normal select:
SELECT id, name FROM users;
and to get the number of entries:
SELECT COUNT(id) FROM users;
Its becuase you are using an aggregate function in the select part of the query,
to return the 10 records you just need the id, and first_name in the query.
EG:
SELECT id, first_Name
FROM users
if you wanted to get a count of the records in the table then you could use
SELECT (Count(id))
FROM [users]
It's not "wrong", but it is meaningless without a "group by" clause - most databases will reject that query, as aggregate functions should include a group by if you're including other columns.
Not sure exactly what you're trying to achieve with this?
select id, first_name,(select count(*) from users) AS usercount from users;
will give each individual user and the total count but again, not sure why you would want it.
select id, first_name from users,(select count(*) as total from users) as t;
COUNT is an aggregate function and it will always give you count of all records in table unless used in combination with group by.
If you use it in combination with normal query, then it will take priority in deciding the final output as in your case it returns 1.
If you want to return all 10 records, you should just write -
select id,first_name from users
If you need number of rows in a table, you can use MySQL's SQL_CALC_FOUND_ROWS clause. Check MySQL docs to see how it's used.
I have a select query like this
select count(distinct id)*100/totalcount as freq, count (distinct id) from
<few joins, conditions, gorup by here> .....
Will this result in 2 calculations of count under MySql 5.0? I can calculate the frequency in my appliation as well if this is a problem. I am aware of the solutions presented in Adding percentages to multiple counts in one SQL SELECT Query but I just want to avoid nested queries
select count(distinct id)*100/totalcount as freq, count (distinct id) from
<few joins, conditions, gorup by here> .....
Yes, it will result in several evaluations.
Each recordset on DISTINCT id will be built separately for each function
Note that if not for DISTINCT, MySQL would use each record only once (though in multiple function calls).
Since COUNT is very cheap, function calls add almost nothing to overall query time.
You can benefit from rewriting your query as this:
SELECT COUNT(id) * 100 / totalcount AS freq,
COUNT(id)
FROM (
SELECT DISTINCT id
FROM original_query
) q
BTW, why do you need both GROUP BY and DISTINCT in one query? Could you please post your original query as it is?