I have a table that looks like this with three columns From, To, and Symbol:
From To Symbol
0 2 dog
2 5 dog
5 9 cat
9 15 cat
15 20 dog
20 40 dog
40 45 dog
I was trying to write an SQL query that groups records in a way that produces the following result:
From To Symbol
0 5 dog
5 15 cat
15 45 dog
That is, if the From and To values are continuous for the same Symbol, one result record is created with the smallest From and the largest To values and the Symbol. In the above example table, since the second record has a value of 2 in the To column which is not the same as the From value in the next record with the same Symbol (15, 20, dog), two result records are created for the same Symbol (dog).
I have tried to join the table to itself, then group by. But I could not figure out how exactly that can be done. I have to do this in Microsoft Access. Any help would be greatly appreciated. Thanks!
Assuming the values have no overlaps and that gaps separate values, you can do this in MS Access with a trick. You need to identify the adjacent symbols that are the same. Well, you can identify them by counting the number of previous rows with different symbols (using a subquery). Once you have this information, the rest is aggregation:
select symbol, min(from) as from, max(to) as to
from (select t.*,
(select count(*)
from t as t2
where t2.from < t.from and t2.symbol <> t.symbol
) as grp
from t
) t
group by symbol, grp;
Gaps would make this problem much harder in MS Access.
Note: Don't use reserved words or keywords for column names. This code uses the names supplied in the question, but doesn't bother to escape them. I think that just makes it harder to understand the query.
Related
i'm looking for a way to, on my query, dynamically set the beginning of the window function on Sql Server using ROWS BETWEEN.
Something like:
SUM(field) OVER(ORDER BY field2 ROWS BETWEEN field3 PRECEDING AND CURRENT ROW)
field3 holds the amount of items (via group by from a CTE) that represent a group.
Is that possible or should i try a different approach?
>> EDIT
My query is too big and messy to share here, but let me try to explain what i need. It's from a report builder which allows users to create custom formulas, like "emplyoees/10". This also allows the user to simply input a formula like "12" and i need to calculate subtotals and the grand total for them. When using a field, like "employees", everything works fine. But for constant values i can't sum the values without rewriting a lot of stuff (which i'm trying to avoid).
So, consider a CTE called "aggregator" and the following query:
SELECT
*,
"employees"/10 as "ten_percent"
12 as "twelve"
FROM aggregator
This query returns this output:
row_type counter company_name department_name employees ten_percent twelve
data 1 A A1 10 1 12
data 1 A A2 15 1,5 12
data 1 A A3 10 1 12
subtotal 3 A 35 3,5 12
data 1 B B1 10 1 12
subtotal 1 B 10 1 12
total 4 45 4,5 12
As you can see, the values fot "twelve" are wrong for subtotal and total row types. I'm trying to solve this without changing the CTE.
ROLLUP won't work because i already have the sum for other columns.
I tried this (i ommited "row_type_sort" on the table above, it defines the sorting):
CASE
WHEN row_type = 'data' THEN
MAX(aggregator.[twelve])
ELSE
SUM(SUM(aggregator.[twelve]))
OVER (ORDER BY "row_type_sort" ROWS BETWEEN unbounded PRECEDING AND CURRENT ROW)
END AS "twelve"
This would work OK if i could change "unbounded" by the value of column "counter", which was my original question.
LAG/LEAD wasn't helpful neither.
I'm out of ideas. Is it possible to achieve what i need only by changing this part of the query, or the result of the CTE should be changed as well?
Thanks
I am pulling data and when I pull in the text field my results for the "distinct ID" are sometimes being duplicated when there are multiple results for that ID. Is there a way to concatenate the results into a single column/row rather than having them duplicated?
It looks like there are ways in other SQL platforms but I have not been able to find something that works in HANA.
Example
Select
Distinct ID
From Table1
If I pull only Distinct ID I get the following:
ID
1
2
3
4
However when I pull the following:
Example
Select
Distinct ID,Text
From Table1
I get something like
ID
Text
1
Dog
2
Cat
2
Dog
3
Fish
4
Bird
4
Horse
I am trying to Concat the Text field when there is more than 1 row for each ID.
What I need the results to be (Having a "break" between results so that they are on separate lines would be even better but at least a "," would work):
ID
Text
1
Dog
2
Cat,Dog
3
Fish
4
Bird,Horse
I see Kiran has just referred to another valid answer in the comment, but in your example this would work.
SELECT ID, STRING_AGG(Text, ',')
FROM TABLE1
GROUP BY ID;
You can replace the ',' with other characters, maybe a '\n' for a line break
I would caution against the approach to concatenate rows in this way, unless you know your data well. There is no effective limit to the rows and length of the string that you will generate, but HANA will have a limit on string length, so consider that.
I use SAP-HANA database. I have a simple 2 column table whose columns are number, name, noodles, fish . The rows are these:
number name noodles fish
1 tom x
1 tom x
1 jack
2 jack x
I would like to group the rows by the id, and concatenate the names into a field, and thus obtain this:
number name noodles fish
1 tom x x
2 jack x
Can you please tell me how we can perform this operation in sap-hana? Thanks in advance.
Well, you did not really concatenate the names, but instead kept the same ones (if you would have concatenated the names as well, you would get something like jackjack in your result). I guess your x's indicate some sort of ABAP-style flags.
In any case, you would do this with grouping. This is a completely non-HANA thing (you can use the same basic SQL for any DB). You can group against several columns. All other columns that you want to select must be used in an aggregated expression (e.g. a SUM, MAX, COUNT, etc.).
To get the output from your question, I wrote the following code:
SELECT "ID", "NAME", MAX("FISH"), MAX("NOODLES")
FROM #TEST GROUP BY "ID", "NAME";
And got the same output as you. I used the MAX function based on the following assumption: you would want to get X if there is any X in the "concatenated" (aggregated) rows in that column. You get nothing / space if all the "concatenated" rows have space in them.
I have a requirement wherein i need to find the record number of the records that are returned from the resultset. I know that i can use ROWNUM to get the record number from the resultset but my issue is slightly different. below are the details
Table : ProcessSummary
Columns:
PS_PK ProcessId StepId AsscoiateId ProcessName AssetAmount
145 25 50 Process1 3,500.00
267 26 45 Process2 4,400.00
356 27 70 Process3 2,400.00
456 28 80 90 Process4 780.00
556 29 56 67 Process5 4,500.00
656 45 70 Process6 6,000.00
789 31 75 Process7 8,000.00
Now what i need to do is fetch all the records from the ProcessSummary Table when either of ProcessId OR StepId OR AssociateId is NULL. I wrote the below query
select * from ProcessSummary where ProcessId IS NULL OR StepId IS NULL OR AsscoiateId IS NULL
As expected i got 1st, 2nd, 3rd, 6th and 7th records in the resultset that got returned.
Now what i need is to get the records numbers 1,2,3,6,7. I tried to use the ROWNUM as below but i got the values of 1,2,3,4,5 and not 1,2,3,6,7.
select ROWNUM from ProcessSummary where ProcessId IS NULL OR StepId IS NULL OR AsscoiateId IS NULL
Is it possible to get the ROWNUM values in the sequence that i want and if yes then can you please let me know how can i do this. Also if ROWNUM cannot be used then what would be the other option that i can use to get the result in the form that i want.
Any help would be greately appericiated as i could not find much on the net or SO regarding this sort of requirement.
Thanks
Vikeng21
rownum is an internal numbering that gives you a row number based on the current query results only, so that numbering is not tied to a specific record, and it will change when you change the data or the query.
But the numbering you ask for is already in your table. It looks like you just need to SELECT PS_PK .. instead. PS_PK is the field in your table that contains the actual number you want.
You can generate a numbering using an analytical function, and then filter that query. You need some fields to order by, though. In this case I've chosen PS_PK, but it can be another field, like ProcessName or a combination of other fields as well.
select
*
from
(select
dense_rank() over (order by PS_PK) as RANKING,
p.*
from
ProcessSummary p)
where
ProcessId IS NULL OR StepId IS NULL OR AsscoiateId IS NULL
So, in this query, first a numbering is calculated for each row that is returned from the inner query. The numbering is returned as the field RANKING. And then the other query filters further, but still will return the field RANKING with the original numbering.
Instead of dense_rank there is also rank and row_number. The differences are subtle, but you can just experiment and read some docs here and here to learn about the differences and see which one fits you best.
Note that this might slow down your query, because the inner query first generates a number for each row in the table (there is no filtering on that level now).
Firstly, I could not think of a better question title. Apologies for that.
So, I am writing a query and here is something(I think) it would return without aggregating functions and group by. I am using this as an example and actual query contains a lot more fields:
SUBJ CLASSROOM CLASSROOM_CAPACITY
A 1 25
B 2 50
C 3 60
A 2 50
A 1 25 <--Not actually duplicate
Now you would say there are duplicate records. But in fact they are not duplicate in a way that there are some extra fields(not shown here) which would have different values for those seemingly duplicate records.
What I want:
SUBJ CLASSROOM CLASSROOM_CAPACITY
A 1 25
2 50
TOTAL 75
B 2 50
TOTAL 50
C 3 60
TOTAL 60
//EDIT - Apparently following line is causing too much confusion. Ignore it. How can I get rest of the table correctly?
TOTAL 135 //It seems its quite difficult to get 135 here. Its ok if this total is messed up
What I am trying:
SELECT
SOME_FIELDS,
SUBJ,
CLASSROOM,
SUM(CLASSROOM_CAPACITY)
FROM
MYTABLE
WHERE .....
GROUP BY SOME_FIELDS, ROLLUP(SUBJ,CLASSROOM)
The problem:
Thanks to those "seemingly duplicate" records, classroom capacities are being summed up multiple times. How do I prevent that? Am I doing this the wrong way?
The actual query is lot more complicated but I think if I can get this right, I can apply it to bigger query.
PS: I know how to get text "Total" instead of blank entry with ROLLUP using GROUPING so you can skip that part.
The cardinality you're introducing is a little off and when you sort the that ROLLUP starts to work. Your saying that:
SUBJ CLASSROOM CLASSROOM_CAPACITY
A 1 25
is equal to:
SUBJ CLASSROOM CLASSROOM_CAPACITY
A 1 25
But the SOME_FIELDS could vary per row. When you aggregate up to just the columns above, what do you expect to happen to SOME_FIELDS?
If these can be ignore for the purposes of this query your best bet is to first find the DISTINCT records (i.e. records that contain a unique tuple of subj, classroom and classroom_capacity) and then do the ROLLUP on this data set. The following query achieves this:
WITH distinct_subj_classrm_capacity AS (
SELECT DISTINCT
subj
, classroom
, classroom_capacity
FROM mytable
)
SELECT
subj
, classroom
, SUM(classroom_capacity)
FROM distinct_subj_classrm_capacity
GROUP BY ROLLUP(subj, classroom)
If you're not interested in the break report results that ROLLUP gives you and you simply want the raw totals then you can use the analytic version of SUM (see here for more on Oracle analytic functions: http://docs.oracle.com/cd/E11882_01/server.112/e26088/functions004.htm)
WITH distinct_subj_classrm_capacity AS (
SELECT DISTINCT
subj
, classroom
, classroom_capacity
FROM mytable
)
SELECT DISTINCT
subj
, SUM(classroom_capacity) OVER (PARTITION BY subj) classroom_capacity_per_subj
FROM distinct_subj_classrm_capacity
This gives results in the format:
SUBJ CLASSROOM_CAPACITY_PER_SUBJ
A 75
B 50
C 60