I have a table and I would like to get a row containing all the latest non-null attributes for each column (without combining separate queries for each column, which doesn't seem elegant to me).
Example:
A B C Time
1 a 7 0
NULL NULL 3 1
3 NULL 4 2
NULL NULL 6 3
Result I seek:
A B C
3 a 6
As I said, I know how to select what I want for each column separately, but I was wondering if there's a better way to do it. No need to tax the poor database if it isn't needed.
Probably a better way than this, but it's Monday and I'm not quite conscious yet:
select #a:=null, #b:=null, #c:=null;
select A,B,C from (
select #a:=coalesce(A,#a) as A, #b:=coalesce(B,#b) as B, #c:=coalesce(C,#) as C time
from yourtable
order by time asc
) as y order by time desc limit 1;
Basically, iterate over each row in the database and build up the "latest" value as you go, then reverse the result set and select only the one with the highest time value
Related
SELECT column_1 FROM table_1,table_2;
When I ran this on my database it returned huge number of rows with duplicate column_1 values. I could not understand why I got these results. Please explain what this query does.
it gives you a cross product from table 1 and table 2
In more layman's terms, it means that for each record in Table A, you get every record from Table B (all possible combinations).
TableA with 3 records and Table B with 3 records gives 9 total records in the result:
TableA-1/B-1
TableA-1/B-2
TableA-1/B-3
TableA-2/B-1
TableA-2/B-2
TableA-2/B-3
TableA-3/B-1
TableA-3/B-2
TableA-3/B-3
Often used as a basis for Cartesian Queries (which themselves are the means to generate, say, a list of future dates based on a recurrence schedule: give me all possible results for the next 6 months, then restrict that set to those whose factor matches my day of the week)
This is 'valid' way of cross joining two tables; it is not the preferred way though. Cross Join would be much clearer. An on condition would then be helpful to limit results,
Imagine that i have 3 friends named Jhon, Ana, Nick; then i have in the other table 2 are T-shirts a red and a yellow and i wanna know witch is from.
So in the query being tableA:Friends and tableB:Tshirts returns:
1|JHON | t-shirt_YELLOW
2|JHON | t-shirt_RED
3|ANA | t-shirt_YELLOW
4|ANA | t-shirt_RED
5|NICK | t-shirt_YELLOW
6|NICK | t-shirt_RED
As you see this join has no relational logic between friends and Tshirts so by evaluating all the posible combination generates what you call duplicates.
I think it will be easiest to start with the table I have and the result I am aiming for.
Name | Date
A | 03/01/2012
A | 03/01/2012
B | 02/01/2012
A | 02/01/2012
B | 02/01/2012
A | 02/01/2012
B | 01/01/2012
B | 01/01/2012
A | 01/01/2012
I want the result of my query to be:
Name | 01/01/2012 | 02/01/2012 | 03/01/2012
A | 1 | 2 | 2
B | 2 | 2 | 0
So basically I want to count the number of rows that have the same date, but for each individual name. So a simple group by of dates won't do because it would merge the names together. And then I want to output a table that shows the counts for each individual date using php.
I've seen answers suggest something like this:
SELECT
NAME,
SUM(CASE WHEN GRADE = 1 THEN 1 ELSE 0 END) AS GRADE1,
SUM(CASE WHEN GRADE = 2 THEN 1 ELSE 0 END) AS GRADE2,
SUM(CASE WHEN GRADE = 3 THEN 1 ELSE 0 END) AS GRADE3
FROM Rodzaj
GROUP BY NAME
so I imagine there would be a way for me to tweak that but I was wondering if there is another way, or is that the most efficient?
I was perhaps thinking if the while loop were to output just one specific name and date each time along with the count, so the first result would be A,01/01/2012,1 then the next A,02/01/2012,2 - A,03/01/2012,3 - B,01/01/2012,2 etc. then perhaps that would be doable through a different technique but not sure if something like that is possible and if it would be efficient.
So I'm basically looking to see if anyone has any ideas that are a bit outside the box for this and how they would compare.
I hope I explained everything well enough and thanks in advance for any help.
You have to include two columns in your GROUP BY:
SELECT name, COUNT(*) AS count
FROM your_table
GROUP BY name, date
This will get the counts of each name -> date combination in row-format. Since you also wanted to include a 0 count if the name didn't have any rows on a certain date, you can use:
SELECT a.name,
b.date,
COUNT(c.name) AS date_count
FROM (SELECT DISTINCT name FROM your_table) a
CROSS JOIN (SELECT DISTINCT date FROM your_table) b
LEFT JOIN your_table c ON a.name = c.name AND
b.date = c.date
GROUP BY a.name,
b.date
SQLFiddle Demo
You're asking for a "pivot". Basically, it is what it is. The real problem with a pivot is that the column names must adapt to the data, which is impossible to do with SQL alone.
Here's how you do it:
SELECT
Name,
SUM(`Date` = '01/01/2012') AS `01/01/2012`,
SUM(`Date` = '02/01/2012') AS `02/01/2012`,
SUM(`Date` = '03/01/2012') AS `03/01/2012`
FROM mytable
GROUP BY Name
Note the cool way you can SUM() a condition in mysql, becasue in mysql true is 1 and false is 0, so summing a condition is equivalent to counting the number of times it's true.
It is not more efficient to use an inner group by first.
Just in case anyone is interested in what was the best method:
Zane's second suggestion was the slowest, I loaded in a third of the data I did for the other two and it took quite a while. Perhaps on smaller tables it would be more efficient, and although I am not working with a huge table roughly 28,000 rows was enough to create significant lag, with the between clause dropping the result to about 4000 rows.
Bohemian's answer gave me the least amount to code, I threw in a loop to create all the case statements and it worked with relative ease. The benefit of this method was the simplicity, besides creating the loop for the cases, the results come in without the need for any php tricks, just simple foreach to get all the columns. Recommended for those not confident with php.
However, I found Zane's first suggestion the quickest performing and despite the need for extra php coding it seems I will be sticking with this method. The disadvantage of this method is that it only gives the dates that actually have data, so creating a table with all the dates becomes a bit more complicated. What I did was create a variable that keeps track of what date it is supposed to be compared to the table column which is reset on each table row, when the result of the query is equal to that date it echoes the value otherwise it does a while loop echoing table cells with 0 until the dates do match. It also had to do a check to see if the 'Name' value is still the same and if not it would switch to the next row after filling in any missing cells with 0 to the end of that row. If anyone is interested in seeing the code you can message me.
Results of the two methods over 3 months of data (a column for each day so roughly 90 case statements) ~ 12,000 rows out of 28,000:Bohemian's Pivot - ~0.158s (highest seen ~0.36s)Zane's Double Group by - ~0.086s (highest seen ~0.15s)
For a dating application, I have a few tables that I need to query for a single output with a LIMIT 10 of both queries combined. It seems difficult to do at the moment, even though it's not an issue to query them separately, but the LIMIT 10 won't work as the numbers are not exact (ex. not LIMIT 5 and LIMIT 5, one query may return 0 rows, while the other 10, depending on the scenario).
members table
member_id | member_name
------------------------
1 Herb
2 Karen
3 Megan
dating_requests
request_id | member1 | member2 | request_time
----------------------------------------------------
1 1 2 2012-12-21 12:51:45
dating_alerts
alert_id | alerter_id | alertee_id | type | alert_time
-------------------------------------------------------
5 3 2 platonic 2012-12-21 10:25:32
dating_alerts_status
status_id | alert_id | alertee_id | viewed | viewed_time
-----------------------------------------------------------
4 5 2 0 0000-00-00 00:00:00
Imagine you are Karen and just logged in, you should see these 2 items:
1. Herb requested a date with you.
2. Megan wants a platonic relationship with you.
In one query with a LIMIT of 10. Instead here are two queries that need to be combined:
1. Herb requested a date with you.
-> query = "SELECT dr.request_id, dr.member1, dr.member2, m.member_name
FROM dating_requests dr
JOIN members m ON dr.member1=m.member_id
WHERE dr.member2=:loggedin_id
ORDER BY dr.request_time LIMIT 5";
2. Megan wants a platonic relationship with you.
-> query = "SELECT da.alert_id, da.alerter_id, da.alertee_id, da.type,
da.alert_time, m.member_name
FROM dating_alerts da
JOIN dating_alerts_status das ON da.alert_id=das.alert_id
AND da.alertee_id=das.alertee_id
JOIN members m ON da.alerter_id=m.member_id
WHERE da.alertee_id=:loggedin_id AND da.type='platonic'
AND das.viewed='0' AND das.viewed_time<da.alert_time
ORDER BY da.alert_time LIMIT 5";
Again, sometimes both tables may be empty, or 1 table may be empty, or both full (where LIMIT 10 kicks in) and ordered by time. Any ideas on how to get a query to perform this task efficiently? Thoughts, advice, chimes, optimizations are welcome.
You can combine multiple queries with UNION, but only if the queries have the same number of columns. Ideally the columns are the same, not only in data type, but also in their semantic meaning; however, MySQL doesn't care about the semantics and will handle differing datatypes by casting up to something more generic - so if necessary you could overload the columns to have different meanings from each table, then determine what meaning is appropriate in your higher level code (although I don't recommend doing it this way).
When the number of columns differs, or when you want to achieve a better/less overloaded alignment of data from two queries, you can insert dummy literal columns into your SELECT statements. For example:
SELECT t.cola, t.colb, NULL, t.colc, NULL FROM t;
You could even have some columns reserved for the first table and others for the second table, such that they are NULL elsewhere (but remember that the column names come from the first query, so you may wish to ensure they're all named there):
SELECT a, b, c, d, NULL AS e, NULL AS f, NULL AS g FROM t1
UNION ALL -- specify ALL because default is DISTINCT, which is wasted here
SELECT NULL, NULL, NULL, NULL, a, b, c FROM t2;
You could try aligning your two queries in this fashion, then combining them with a UNION operator; by applying LIMIT to the UNION, you're close to achieving your goal:
(SELECT ...)
UNION
(SELECT ...)
LIMIT 10;
The only issue that remains is that, as presented above, 10 or more records from the first table will "push out" any records from the second. However, we can utilise an ORDER BY in the outer query to solve this.
Putting it all together:
(
SELECT
dr.request_time AS event_time, m.member_name, -- shared columns
dr.request_id, dr.member1, dr.member2, -- request-only columns
NULL AS alert_id, NULL AS alerter_id, -- alert-only columns
NULL AS alertee_id, NULL AS type
FROM dating_requests dr JOIN members m ON dr.member1=m.member_id
WHERE dr.member2=:loggedin_id
ORDER BY event_time LIMIT 10 -- save ourselves performing excessive UNION
) UNION ALL (
SELECT
da.alert_time AS event_time, m.member_name, -- shared columns
NULL, NULL, NULL, -- request-only columns
da.alert_id, da.alerter_id, da.alertee_id, da.type -- alert-only columns
FROM
dating_alerts da
JOIN dating_alerts_status das USING (alert_id, alertee_id)
JOIN members m ON da.alerter_id=m.member_id
WHERE
da.alertee_id=:loggedin_id
AND da.type='platonic'
AND das.viewed='0'
AND das.viewed_time<da.alert_time
ORDER BY event_time LIMIT 10 -- save ourselves performing excessive UNION
)
ORDER BY event_time
LIMIT 10;
Of course, now it's up to you to determine what type of row you're dealing with as you read each record in the resultset (suggest you test request_id and/or alert_id for NULL values; alternatively one could add an additional column to the results that explicitly states from which table each record originated, but it should be equivalent provided those id columns are NOT NULL).
Is it possible to sort in MySQL by "order by" using a predefined set of column values (ID) like order by (ID=1,5,4,3) so I would get records 1, 5, 4, 3 in that order out?
UPDATE: Why I need this...
I want my records to change sort randomly every 5 minutes. I have a cron task to update the table to put different, random sort order in it.
There is just one problem! PAGINATION.
I will have visitors who come to my page, and I will give them the first 20 results. They will wait 6 minutes, go to page 2 and have the wrong results as the sort order has already changed.
So I thought that if I put all the IDs into a session on page 2, we get the correct records even if the sorting had already changed.
Is there any other better way to do this?
You can use ORDER BY and FIELD function.
See http://lists.mysql.com/mysql/209784
SELECT * FROM table ORDER BY FIELD(ID,1,5,4,3)
It uses Field() function, Which "Returns the index (position) of str in the str1, str2, str3, ... list. Returns 0 if str is not found" according to the documentation. So actually you sort the result set by the return value of this function which is the index of the field value in the given set.
You should be able to use CASE for this:
ORDER BY CASE id
WHEN 1 THEN 1
WHEN 5 THEN 2
WHEN 4 THEN 3
WHEN 3 THEN 4
ELSE 5
END
On the official documentation for mysql about ORDER BY, someone has posted that you can use FIELD for this matter, like this:
SELECT * FROM table ORDER BY FIELD(id,1,5,4,3)
This is untested code that in theory should work.
SELECT * FROM table ORDER BY id='8' DESC, id='5' DESC, id='4' DESC, id='3' DESC
If I had 10 registries for example, this way the ID 1, 5, 4 and 3 will appears first, the others registries will appears next.
Normal exibition
1
2
3
4
5
6
7
8
9
10
With this way
8
5
4
3
1
2
6
7
9
10
There's another way to solve this. Add a separate table, something like this:
CREATE TABLE `new_order` (
`my_order` BIGINT(20) UNSIGNED NOT NULL,
`my_number` BIGINT(20) NOT NULL,
PRIMARY KEY (`my_order`),
UNIQUE KEY `my_number` (`my_number`)
) ENGINE=INNODB;
This table will now be used to define your own order mechanism.
Add your values in there:
my_order | my_number
---------+----------
1 | 1
2 | 5
3 | 4
4 | 3
...and then modify your SQL statement while joining this new table.
SELECT *
FROM your_table AS T1
INNER JOIN new_order AS T2 on T1.id = T2.my_number
WHERE ....whatever...
ORDER BY T2.my_order;
This solution is slightly more complex than other solutions, but using this you don't have to change your SELECT-statement whenever your order criteriums change - just change the data in the order table.
If you need to order a single id first in the result, use the id.
select id,name
from products
order by case when id=5 then -1 else id end
If you need to start with a sequence of multiple ids, specify a collection, similar to what you would use with an IN statement.
select id,name
from products
order by case when id in (30,20,10) then -1 else id end,id
If you want to order a single id last in the result, use the order by the case. (Eg: you want "other" option in last and all city list show in alphabetical order.)
select id,city
from city
order by case
when id = 2 then city else -1
end, city ASC
If i had 5 city for example, i want to show the city in alphabetical order with "other" option display last in the dropdown then we can use this query.
see example other are showing in my table at second id(id:2) so i am using "when id = 2" in above query.
record in DB table:
Bangalore - id:1
Other - id:2
Mumbai - id:3
Pune - id:4
Ambala - id:5
my output:
Ambala
Bangalore
Mumbai
Pune
Other
SELECT * FROM TABLE ORDER BY (columnname,1,2) ASC OR DESC
I want to have a query that returns the best results from a table.
I am defining the best results to be the addition of two columns a + b (each column holds an int)
ie:
entry a b
1 4 5
2 3 2
3 20 30
Entry 3 would be returned because a + b is the highest in this case.
Is there a way to do this? One idea I had was to create another column in the table which holds the addition of a and b and then ORDER by DESC, but that seems a little bit messy.
Any ideas?
Thanks!
SELECT *
FROM mytable
ORDER BY
a + b DESC
LIMIT 1
Adding another column, however, would be a good option, since you could index this column which would improve the query.