I have a table with this sort of data:
+------------+----------+----------+
| Unique ID | Name | Class |
+------------+----------+----------+
| 1 | Name 1 | Class A |
| 2 | Name 2 | "" |
| 3 | Name 3 | Class C |
| 4 | Name 1 | "" |
| 5 | Name 4 | "" |
| 6 | Name 4 | "" |
+------------+----------+----------+
I am trying to do something I thought was simple, but i did not find so.
I am trying to "extract" only the lines with an empty string value in 'Class' for a group of equal names.
So in this case I would get this result :
+------------+----------+--------+
| Unique ID | Name | Class |
+------------+----------+--------+
| 2 | Name 2 | "" |
| 5 | Name 4 | "" |
+------------+----------+--------+
So not Name 1 because even though there is a line with "" there is another line with 'Class A'.
I thought a UNION would do the job but I am not gettgin anything because I think unions are for two tables but the problem here is I have the data in the same table.
Thank you for your help
Access syntax may be a bit different but this returns what you want in Oracle:
SELECT distinct Name, Class FROM table1 Where Name NOT in (select name from table1 where class is not null)
A Union melds two result sets, whether or not they come from the same table is irrelevant. What you want to do is omit from the result set the rows with the same name AND class is not null. Not having your query to expand on or change is a problem, but if you add a clause that says something like where "name not in (select name from table where class is not null);", that may do it.
Related
I have two tables in a MySQL database like this:
User:
userid |userid | Username | Plan(VARCHAR) | Status |
-----------+------------+--------------+---------------+---------+
1 | 1 | John | 1,2,3 |1 |
2 | 2 | Cynthia | 1,2 |1 |
3 | 3 | Charles | 2,3,4 |1 |
Plan: (planid is primary key)
planid(INT) | Plan_Name | Cost | status |
-------------+----------------+----------+--------------+
1 | Tamil Pack | 100 | ACTIVE |
2 | English Pack | 100 | ACTIVE |
3 | SportsPack | 100 | ACTIVE |
4 | KidsPack | 100 | ACTIVE |
OUTPUT
id |userid | Username | Plan | Planname |
---+-------+----------+------------+-------------------------------------+
1 | 1 | John | 1,2,3 |Tamil Pack,English Pack,SportsPack |
2 | 2 | Cynthia | 1,2 |Tamil Pack,English Pack |
3 | 3 | Charles | 2,3,4 |English Pack,Sportspack, Kidspack |
Since plan id in Plan table is integer and the user can hold many plans, its stored as comma separated as varchar, so when i try with IN condition its not working.
SELECT * FROM plan WHERE find_in_set(plan_id,(select user.planid from user where user.userid=1))
This get me the 3 rows from plan table but i want the desired output as above.
How to do that.? any help Please
A rewrite off your query what should work is as follows..
Query
SELECT
all columns you need
, GROUP_CONCAT(Plan.Plan_Name ORDER BY Plan.planid) AS Planname
FROM
Plan
WHERE
FIND_IN_SET(Plan.plan_id,(
SELECT
User.Plan
FROM
user
WHERE User.userid = 1
)
)
GROUP BY
all columns what are in the select (NOT the GROUP_CONCAT function)
You also can use FIND_IN_SET on the ON clause off a INNER JOIN.
One problem is that the join won't ever use indexes.
Query
SELECT
all columns you need
, GROUP_CONCAT(Plan.Plan_Name ORDER BY Plan.planid) AS Planname
FROM
User
INNER JOIN
Plan
ON
FIND_IN_SET(Plan.id, User.Plan)
WHERE
User.id = 1
GROUP BY
all columns what are in the select (NOT the GROUP_CONCAT function)
Like i said in the comments you should normalize the table structures and add the table User_Plan whats holds the relations between the table User and Plan.
I haven't touched the backend in a while.. so forgive me if this is super simple. I'm working with Lumen v.5.6.1.
| table.sets | | table.indexed_items |
|----------------| |---------------------------------|
| ID | SET | | ID | setId | itemId | have |
|----|-----------| |----|-------|--------|-----------|
| 1 | set name 1| | 1 | 3 | 1 | 2 |
| 2 | set name 2| | 2 | 3 | 2 | 1 |
| 3 | set name 3| | 3 | 3 | 3 | 4 |
| 4 | 2 | 4 | 1 |
| 5 | 2 | 5 | 3 |
| 6 | 2 | 6 | 1 |
How would I return in one query, groupedBy/distinct by setId (with set name as a left join?) to have a return like this:
[
setId: 2,
name: 'set name 2',
haveTotal: 5,
],
[
setId: 3,
name: 'set name 3',
haveTotal: 7,
]
Here is a raw MySQL query which should work. To convert this to Laravel should not be too much work, though you might need to use DB::raw once or twice.
SELECT
s.ID AS setId,
s.`SET` AS name,
COALESCE(SUM(ii.have), 0) AS haveTotal
FROM sets s
LEFT JOIN indexed_items ii
ON s.ID = ii.setId
GROUP BY
s.ID;
Demo
If you don't want to return sets having no entries in the indexed_items table, then you may remove the call to COALESCE, and you may also use an inner join instead of a left join.
Note that using SET to name your tables and columns is not a good idea because it is a MySQL keyword.
If you are using or want to use eloquent, you can do something like:
$sets = App\Sets::withCount('indexed_items')->get();
This will return a collection with a column name indexed_items_count
Obviously you will need to change depending on your model names.
Here are the docs
I always use in my project for count relation ship record.
$sets->indexed_items->count();
I'm currently trying to update an existing database (removing duplicates).
You can see the structure as follows :
I have a database on which specific entries are marked as "Main". These entries need to be updated with data from duplicate records, only having the same name.
(Updated table to reflect my question better)
It would look like this:
+----+------+-----------------+--------------+---------+
| ID | Name | Field-To-Update | Duplication | Source |
+----+------+-----------------+--------------+---------+
| . | A | xxx | Main | 1 |
| . | A | yyy | "" | 2 |
| . | A | zzz | "" | 3 |
| . | B | foo | "" | 1 |
| . | B | bar | Main | 2 |
+----+------+-----------------+--------------+---------+
Should result in
+----+------+-----------------+--------------+-----------------------------------------------+
| ID | Name | Field-To-Update | Duplication | Source |
+----+------+-----------------+--------------+-----------------------------------------------+
| . | A | yyy | Main | 1 |
| . | A | yyy | "" | 2 |
| . | A | zzz | "" | 3 (should be updated from a specific source) |
| . | B | bar | "" | 1 |
| . | B | bar | Main | 2 (should be updated from a specific source) |
+----+------+-----------------+--------------+-----------------------------------------------+
Do any of you have an idea how to tackle this? I've tried with multiple queries for a couple of days now without any success.
you could use a update with join
update t
set t.field_to_update = x.field_to_update
from your_table t
inner join ( select name, field_to_update
from your_table
where Duplication <> 'Main') x
) on t.name = x.name
where t.Duplication = 'Main'
Bizarre requirement. You say you are updating to remove duplicates, however, it looks to me like you are creating duplicates.
There are only 2 records for each Name? Try:
UPDATE Table INNER JOIN
(SELECT Name, [Field-To-Update]
FROM Table
WHERE Duplication Is Null) AS Query1 ON
Table.Name = Query1.Name SET Table.[Field-To-Update] = [Query1]![Field-To-Update] WHERE Duplication = "Main";
Name is a reserved word in Access. Should not use reserved words as name for anything.
Darn! Did not see #scaisEdge answer before posting.
I tried the <> "Main" criteria and it did not work which was surprising - no records returned. So I switched to the Is Null parameter.
If you want to pull value from the maximum Source for each Name, then need a query that does that. Review http://allenbrowne.com/subquery-01.html#TopN. Then use that query in the above example as Query1.
Hello i want to split a resulting column in multiple columns just like on the link. But number of columns are not specific ;
Example
COL1 | OTHER COLUMNS
----------------------------------------
this,will,split | some value
also,this | some value
this,is,four,columns | some value
I want make this something like that ;
COL1 | COL2 | COL3 | COL4 | OTHER
----------------------------------------
this | will | split| NULL | some value
also | this | NULL | NULL | some value
this | is | four | columns| some value
edit
it looks like similar that question but not:
Can you split/explode a field in a MySQL query?
I want results in 1 row, I dont want something like that;
RESULT
-----
this
will
split
...
on that question you can see there is specific number of cols. bu i dont. :(
How to split a resulting column in multiple columns
I think you can create one relational table and add multiple entry in relational table, hear you don't need to think about column, you have to add entry in row.
eg.
Table 1:
ID | COL1 | OTHER COLUMNS
----------------------------------------
1 |this,will,split | some value
2 |also,this | some value
3 |this,is,four,columns | some value
Table2
ID | Table1_id | value
-------------------------
1 | 1 | this
2 | 1 | will
3 | 1 | split
4 | 2 | also
5 | 2 | this
6 | 3 | this
6 | 3 | is
6 | 3 | four
6 | 3 | columns
Please check this, i think fix your problem.
I'm wondering how to filter Hibernate results.
For example, I have this example table:
--------------------------------------------
| ID | STRING | DATE |
--------------------------------------------
| 1 | "ABC" | 2014-11-07 21:45:00 |
--------------------------------------------
| 2 | "ABC" | 2014-11-07 22:45:00 |
--------------------------------------------
| 3 | "DCE" | 2014-11-07 22:48:00 |
--------------------------------------------
| 4 | "ABC" | 2014-11-07 23:48:00 |
--------------------------------------------
The result that I need is:
--------------------------------------------
| ID | STRING | DATE |
--------------------------------------------
| 3 | "DCE" | 2014-11-07 22:48:00 |
--------------------------------------------
| 4 | "ABC" | 2014-11-07 23:48:00 |
--------------------------------------------
All lines of table, BUT, when I have the same content in column STRING, the column with the most recent DATE is the one in result set.
In hibernate I know
- select ALL lines;
- select the repeated lines and keep only the most recent.
But I don't know how to combine the two situations together, and I can't find it referenced anywhere.
I have a fixed length in STRING, and I use a fake primary key in my code to identify the register, but the value is inserted by user, and I have no previous knowledge of the content in the column.
You can do this using a not exists clause:
select e.*
from example e
where not exists (select 1 from example e2 where e2.string = e.string and e2.date > e.date);
That is, select all rows where there is no other row with the same string and a larger date.