I'd like to consult one thing. I have table in DB. It has 2 columns and looks like this:
Name...bilance
Jane...+3
Jane...-5
Jane...0
Jane...-8
Jane...-2
Paul...-1
Paul...2
Paul....9
Paul...1
...
I have to walk through this table and if I find record with different "name" (than was on previous row) I process all rows with the previous "name". (If I step on the first Paul row I process all Jane rows)
The processing goes like this:
Now I work only with Jane records and walk through them one by one. On each record I stop and compare it with all previous Jane rows one by one.
The task is to sumarize "bilance" column (in the scope of actual person) if they have different signs
Summary:
I loop through this table in 3 levels paralelly (nested loops)
1st level = search for changes of "name" column
2nd level = if change was found, get all rows with previous "name" and walk through them
3rd level = on each row stop and walk through all previous rows with current "name"
Can this be solved only using CURSOR and FETCHING, or is there some smoother solution?
My real table has 30 000 rows and 1500 people and If I do the logic in PHP, it takes long minutes and than timeouts. So I would like to rewrite it to MS SQL 2000 (no other DB is allowed). Are cursors fast solution or is it better to use something else?
Thank you for your opinions.
UPDATE:
There are lots of questions about my "summarization". Problem is a little bit more difficult than I explained. I simplified it just to describe my algorithm.
Each row of my table contains much more columns. The most important is month. That's why there are more rows for each person. Each is for different month.
"Bilances" are "working overtimes" and "arrear hours" of workers. And I need to sumarize + and - bilances to neutralize them using values from previous months. I want to have as many zeroes as possible. All the table must stay as it is, just bilances must be changed to zeroes.
Example:
Row (Jane -5) will be summarized with row (Jane +3). Instead of 3 I will get 0 and instead of -5 I will get -2. Because I used this -5 to reduce +3.
Next row (Jane 0) won't be affected
Next row (Jane -8) can not be used, because all previous bilances are negative
etc.
You can sum all the values per name using a single SQL statement:
select
name,
sum(bilance) as bilance_sum
from
my_table
group by
name
order by
name
On the face of it, it sounds like this should do what you want:
select Name, sum(bilance)
from table
group by Name
order by Name
If not, you might need to elaborate on how the Names are sorted and what you mean by "summarize".
I'm not sure what you mean by this line... "The task is to sumarize "bilance" column (in the scope of actual person) if they have different signs".
But, it may be possible to use a group by query to get a lot of what you need.
select name, case when bilance < 0 then 'negative' when bilance >= 0 then 'positive', count(*)
from table
group by name, bilance
That might not be perfect syntax for the case statement, but it should get you really close.
Related
Here I have this query that finds out the drop percentage of a bunch of clients based on the orders they have received(i.e. It finds the percentage difference in orders by comparing the current month with the previous month). What I want to achieve here is to have a field where I can see the clients who had 4 months continuous drop, 3 months drop, 2 months drop, and 1 month drop.
I know, it can only be achieved by comparing the last 4 months using the lag function or sub queries. can you guys pls help me out on this one, would appreciate it very much
select
fd.customers2, fd.Month1, fd.year1, fd.variance, case when
(fd.variance < -0.00001 and fd.year1 = '2022.0' and fd.Month1 = '1')
then '1month drop' else fd.customers2 end as 1_most_host_drop
from
(SELECT
c.*,
sa.customers as customers2,
sum(sa.order) as orders,
date_part(mon, sa.date) as Month1,
date_part(year, sa.date) as year1,
(cast(orders - LAG(orders) OVER(Partition by customers2 ORDER BY
year1, Month1) as NUMERIC(10,2))/NULLIF(LAG(orders)
OVER(partition by customers2 ORDER BY year1, Month1) * 1, 0)) AS variance
FROM stats sa join (select distinct
d.id, d.customers
from configer d
) c on sa.customers=c.customers
WHERE sa.date >= '2021-04-1'
GROUP BY Month1, sa.customers, c.id, year1,
c.customers)fd
In a spirit of friendliness: I think you are a little premature in posting this here as there are several issues with the syntax before even reaching the point where you can solve the problem:
You have at least two places with a comma immediately preceding the word FROM:
...AS variance, FROM stats_archive sa ...
...d.customers, FROM config d...
Recommend you don't use VARIANCE as an alias (it is a system function in PostgreSQL and so is likely also a system function name in Redshift)
Not super important, but there's no need for c.* - just select the columns you will use
DATE_PART requires a string as the first parameter DATE_PART('mon',current_date)
I might be wrong about this, but I suspect you cannot use column aliases in the partition by or order by of a window function. Put the originating expressions there instead:
... OVER (PARTITION BY customers2 ORDER BY DATE_PART('year',sa.date),DATE_PART('mon',sa.date))
LAG has three parameters. (1) The column you want to retrieve the value from, (2) the row offset, where a positive integer indicates how many rows prior to the current row you should retrieve a value from according to the partition and order context and (3) the value the function should return as a default (in case of the first row in the partition). As such, you don't need NULLIF. So, to get the row immediately prior to the current row, or return 0 in case the current row is the first row in the partition:
LAG(orders,1,0) OVER (PARTITION BY customers2 ORDER BY DATE_PART('year',sa.date),DATE_PART('mon',sa.date))
If you use 0 as a default in the calculation of what is currently aliased variance, you will almost certainly run into a div/0 error either now, or worse, when you least expect it in the future. You should protect against that with some CASE logic or better, provide a more appropriate default value or even better, calculate the LAG with the default 0, then filter out the 0 rows before doing the calculation.
You can't use column aliases in the GROUP BY. You must reference each field that is not participating in an aggregate in the group by, whether through direct mention (sa.date) or indirectly in an expression (DATE_PART('mon',sa.date))
Your date should be '2021-04-01'
All in all, without sample data, expected results using the posted sample data and without first removing syntax errors, it is a tall order to attempt to offer advice on the problem which is any more specific than:
Build the source of the calculation as a completely separate query first. Calculate the LAG in that source query. Only when you've run that source query and verified that the LAG is producing the correct result should you then wrap it as a sub-query or CTE (not sure if Redshift supports these, but presumably) at which point you can filter out the rows with a zero as the denominator (the first month of orders for each customer).
Good luck!
I have the following MYSQL table:
measuredata:
- ID (bigint)
- timestamp
- entityid
- value (double)
The table contains >1 billion entries. I want to be able to visualize any time-window. The time window can be size of "one day" to "many years". There are measurement values round about every minute in DB.
So the number of entries for a time-window can be quite different. Say from few hundrets to several thousands or millions.
Those values are ment to be visualiuzed in a graphical chart-diagram on a webpage.
If the chart is - lets say - 800px wide, it does not make sense to get thousands of rows from database if time-window is quite big. I cannot show more than 800 values on this chart anyhow.
So, is there a way to reduce the resultset directly on DB-side?
I know "average" and "sum" etc. as aggregate function. But how can I i.e. aggregate 100k rows from a big time-window to lets say 800 final rows?
Just getting those 100k rows and let the chart do the magic is not the preferred option. Transfer-size is one reason why this is not an option.
Isn't there something on DB side I can use?
Something like avg() to shrink X rows to Y averaged rows?
Or a simple magic to just skip every #th row to shrink X to Y?
update:
Although I'm using MySQL right now, I'm not tied to this. If PostgreSQL f.i. provides a feature that could solve the issue, I'm willing to switch DB.
update2:
I maybe found a possible solution: https://mike.depalatis.net/blog/postgres-time-series-database.html
See section "Data aggregation".
The key is not to use a unixtimestamp but a date and "trunc" it, avergage the values and group by the trunc'ed date. Could work for me, but would require a rework of my table structure. Hmm... maybe there's more ... still researching ...
update3:
Inspired by update 2, I came up with this query:
SELECT (`timestamp` - (`timestamp` % 86400)) as aggtimestamp, `entity`, `value` FROM `measuredata` WHERE `entity` = 38 AND timestamp > UNIX_TIMESTAMP('2019-01-25') group by aggtimestamp
Works, but my DB/index/structue seems not really optimized for this: Query for last year took ~75sec (slow test machine) but finally got only a one value per day. This can be combined with avg(value), but this further increases query time... (~82sec). I will see if it's possible to further optimize this. But I now have an idea how "downsampling" data works, especially with aggregation in combination with "group by".
There is probably no efficient way to do this. But, if you want, you can break the rows into equal sized groups and then fetch, say, the first row from each group. Here is one method:
select md.*
from (select md.*,
row_number() over (partition by tile order by timestamp) as seqnum
from (select md.*, ntile(800) over (order by timestamp) as tile
from measuredata md
where . . . -- your filtering conditions here
) md
) md
where seqnum = 1;
I am trying to count the amount of urls we have in field in sql I have googled but cannot find anything !
So for example this could be in field "url" row 1 / id 1
url/32432
url/32434
So for example this could be field "url" in row 2 / id 2
url/32432
url/32488
url/32477
So if you were to run the query the count would be 5. There is no comma in between them, only space.
Kind Regards
Scott
This is a very bad layout for data. If you have multiple urls per id, then they should be stored as separate rows in another table.
But, sometimes we are stuck with other people's bad design decisions. You can do something like this:
select (length(replace(urls, 'url', 'urlx')) - length(urls)) as num_urls
Note that the specific functions for length() and replace() might vary, depending on the database.
I have the following query that is part of a common table expression. I don't understand the function of the "Select -1" statement. It is obviously different than the "Select 1" that is used in "EXISTS" statements. Any ideas?
select days_old,
count(express_cd),
count(*),
case
when round(count(express_cd)*100.0/count(*),2) < 1 then '0'
else ''
end ||
cast(decimal(round(count(express_cd)*100.0/count(*),2),5,2) as varchar(7)) ||
'%'
from foo.bar
group by days_old
union all
select -1, -- Selecting the -1 here
count(express_cd),
count(*),
case
when round(count(express_cd)*100.0/count(*),2) < 1 then '0'
else ''
end ||
cast(decimal(round(count(express_cd)*100.0/count(*),2),5,2) as varchar(7)) ||
'%'
from foo.bar
where days_old between 1 and 7
It's just selecting the number "minus one" for each row returned, just like "select 1" will select the number "one" for each row returned.
There is nothing special about the "select 1" syntax uses in EXISTS statements by the way; it's just selecting some random value because EXISTS requires a record to be returned and a record needs data; the number 1 is sufficient.
Why you would do this, I have no idea.
When you have a union statement, each part of the union must contain the same columns. From what I read when I look at this, the first statement is giving you one line for each days old value and then some stats for each day old. The second part of the union is giving you a summary of all the records that are only a week or so less. Since days old column is not relevant here, they put in a fake value as a placeholder in order to do the union. OF course this is just a guess based on reading thousands of queries through the years. To be sure, I would need to actually run teh code.
Since you say this is a CTE, to really understand why this is is happening, you may need to look at the data it generates and how that data is used in the next query that uses the CTE. That might answer your question.
What you have asked is basically about a business rule unique to your company. The true answer should lie in any requirements documents for the original creation of the code. You should go look for them and read them. We can make guesses based on our own experience but only people in your company can answer the why question here.
If you can't find the documentation, then you need to talk (Yes directly talk, preferably in person) to the Stakeholders who use the data and find out what their needs were. Only do this after running the code and analyzing the results to better understand the meaning of the data returned.
Based on your query, all the records with days_old between 1 and 7 will be output as '-1', that is what select -1 does, nothing special here and there is no difference between select -1 and select 1 in exists, both will output the records as either 1 or -1, they are doing the same thing to check whether if there has any data.
Back to your query, I noticed that you have a union all and compare each four columns you select connected by union all, I am guessing your task is to get a final result with days_old not between 1 and 7 and combine the result with day_old, which is one because you take all between 1 and 7.
It is just a grouping logic there.
Your query returns aggregated
data (counts and rounds) grouped by days_old column plus one more group for data where days_old between 1 and 7.
So, -1 is just another additional group there, it cannot be 1 because days_old=1 is an another valid group.
result will be like this:
row1: days_old=1 count(*)=2 ...
row2: days_old=3 count(*)=5 ...
row3: days_old=9 count(*)=6 ...
row4: days_old=-1 count(*)=7
I am looking for a way to search for a certain number of rows as a quality check. For example, we have tables that have a certain set of results that are needed.
Here is a quick table for an example:
ID: Name: Result: Reportable:
ONE A 10 X
TWO B 12 X
THREE C 1
FOUR D 18 X
FOUR(redo) D 11 X
So we are looking to double check results as there are people who accidentally report results multiple times (as in the case with ID FOUR). We have used having counts but we need the numbers to be specific and need a query to verify that number is satisfied.
In the table above we only want IDs ONE, TWO, and FOUR, however we have 4 results (one extra). Currently we have our check showing the count needed (ie 3) and the current result count (4) to show the mismatch but want a query to easily only show the result needed. We would need the redo result most of the time so we have set it so we take the latest date, but it doesn't help filter how many rows or results. I apologize if anything is confusing and I am not able to share the SQL query that we have currently. It's my first time posting so if I need to clarify anything please let me know as this seems to be very complicated. Thank you for your time.
EDIT: The details
We have one table (Table A) letting us know which results are reportable. The ones that are reportable go into another table (Table B). We have had issues in which people have made too many results reportable which overpopulates the Table B. Our old query had a count in Table B, but due to mistakes in people placing multiple reportables, samples which had many redos seem to be finished as they were all placed and met the count in Table B.
So now by using the Table A that helps tell us how many are Reportable, we want this to double check that the samples are indeed ready.
As I understand the question, you want ids that have multiple reportables. Assuming you really mean name, then:
select name
from t
where reportable = 'X'
group by name
having count(*) >= 2;