my sql knowledge is fairly basic and I would be grateful for some advice. I have a table with columns like:
date, time, readings, .... comments1, comments2
What I would like to do is filter the table to show the results when comments1 is equal to a string, which I can achieve. The tricky bit is I then want to find the readings when the time is between 5 and 7 hours after the times returned/identified by the initial query (comments1 = string"). Is there a way to do this with and what would be the best strategy?
Thank you.
You should really store date and time in a single column, otherwise midnight boundaries are extremely difficult to select across. My example assumes your "date" column is a datetime type that also stores the timestamp.
I believe something like this is what you're looking for:
WITH CommentTime AS (
SELECT TOP 1 date
FROM tblRecords
WHERE comments1 = 'The comment to find'
)
SELECT *
FROM tblRecords
WHERE date >= DATEADD(hour, 5, (SELECT date FROM CommentTime))
AND date < DATEADD(hour, 7, (SELECT date FROM CommentTime))
New Answer:
(Reworking Dans answer to instead use a variable)
DECLARE #CommentTime AS DateTime = (SELECT TOP 1 [date] FROM tblRecords WHERE comments1 = 'string')
SELECT * FROM tblRecords
WHERE [date] >= DATEADD(HOUR, 5, #CommentTime)
AND [date] < DATEADD(HOUR, 7, #CommentTime)
Related
Suppose you have a room which is 100sqft and you want to rent it from 1st Aug to 31st Aug.
Bookings Table schema
startdate|enddate|area|storageid
you have following bookings
06-Aug|25-Aug|50|'abc'
05-Aug|11-Aug|40|'xyz'
18-Aug|23-Aug|30|'pqr'
13-Aug|16-Aug|10|'qwe'
Now somebody requests for booking from 08-Aug to 20-Aug. For this date range the maximum area available is 10sqft (Since, for dates 8,9,10 and 11 Aug only 10sq ft is available.)
How would you create an efficient SQL query to get this? Right now I have very messy and inefficient query which gives wrong results for some cases. I am not posting the query because It is so messy that I can't explain it myself.
I don't necessarily want to solve it using SQL only. If there is an algorithm that can solve it efficiently I would extract all the data from database.
Someone removed SQL Server, but here is the algorithm:
DECLARE #startDate date = '2016-08-09';
DECLARE #endDate date = '2016-08-20';
DECLARE #totalArea decimal(19,2) = 100;
WITH Src AS --Your source table
(
SELECT * FROM (VALUES
('2016-08-06', '2016-08-25', 50, 'abc'),
('2016-08-05', '2016-08-11', 40, 'xyz'),
('2016-08-18','2016-08-23',30,'pqr'),
('2016-08-13','2016-08-16',10,'qwe')
)T(startdate, enddate, area, storageid)
), Nums AS --Numbers table 0..N, N must be greater than ranges calculated
(
SELECT 0 N
UNION ALL
SELECT N+1 N FROM Nums
WHERE N<DATEDIFF(DAY,#startDate,#endDate)
) --Query
--You can use total-maxUsed from range of days
SELECT #totalArea-MAX(Used) FROM
(
--Group by day, sum all used areas
SELECT MidDate, SUM(Used) Used FROM
(
--Join table with numbers, split every day, if room used, return area
SELECT DATEADD(DAY, N, #startDate) MidDate, CASE WHEN DATEADD(DAY, N, #startDate) BETWEEN startDate AND endDate THEN area END Used
FROM Src
CROSS APPLY Nums
) T
GROUP BY MidDate
) T
im having a problem where i cant think of a solution, maybe im having a bad table-structure or i just dont know enough about mysql select commands to think of a good solution. Maybe you can help me out:
So i got a table that has a Column with the Date-format (yyyy-mm-dd) i wanted to select all upcoming dates so i did:
SELECT * WHERE date >= now.
This worked kinda well but i also got "dates" where only the year is entered (2014-00-00) i also wanted to select these but "now" is already bigger so i made another column with the year only and if the month, date or both arent known i will use 0000-00-00 and the Column "year" now i could select like this:
SELECT * WHERE date >= now AND year >=now(year)
Now all entrys with 0000-00-00 wont be selected. If i use OR the entrys from last year will be shown.
So thats my problem, is there any way i can change my table so i can have entries with only the year or only year and month and of course all together? I already considered get rid of the date-format and use simple INT with seperated columns for year, month and date. But i think i will have the same problem.
Sometimes i just want to do a capsuled select like
SELECT *
WHERE (date >= now AND year >= now(year))
OR date == "0000-00-00" (i know that this doesnt work)
If I understood your problem correctly, you could use this request:
WHERE (date >= now OR year > now(year))
There is probably a simpler way though, that would preserve your design, like initializing at January 1st (01-01) instead of 00-00
I think you can use this code:
$_SESSION['month'] = //set here your selected month
$_SESSION['year'] = //set here your selected year
SELECT * FROM table WHERE DATEPART(m,date) >= '".$_SESSION['month']."' AND DATEPART(yyyy,year) >= '".$_SESSION['year']."' AND date <> '0000-00-00'
Change your table structure format. Actually just allow for that field to have null value when not entered. By default it will be null then. You shouldn't be storing 0000-00-00 as a value for Date type field. I would rather leave it as null , or as suggested in some of previous answers, initialize it with some other date. It would be much easier to manipulate with database then.
the problem is that half of you write is not MySQL and your database schema is terrible...
You have the following problems:
column data date does not have the date data type.
To fix it, you need to add a cast to the select statement eg. cast(datecolumn as date)
select * from table where cast(datecolumn as date) >= '2014-01-10';
the way to use now date is using the now function.
select now(), date(now());
result> 2014-01-10 11:11:36, 2014-01-10
select * from table where cast(datecolumn as date) >= date(now());
Because your datecolumn is not a date (2014-00-00 is not a valid date), you need to use string manipulation to extract the year.
select substring('2014-01-01', 1,4)
result> 2014
select * from table where substring(datecolumn, 1,4) = year(now());
The comparassion operator is = and not ==
the select statement syntax looks like this (pay attention because you are missing the table in your statement)
select * from [Table] where [column] = condition ...
You probably need or instead of ands, therefore your query should look like this:
select * from FooTable where
cast(datecolumn as date) >= date(now())
or substring(datecolumn, 1,4) >= year(now())
or datecolumn = '0000-00-00'
You should use something like phpmyAdmin or mySQL workbench to test your sql queries before try to use them on php, java or whatever is your programing language.
I am in the situation where i want to match a date range with another date range, it might be simple but i am stuck at it.
Below is table structure
Table - lifecycles
life_id
life_start_date
life_end_date
then a few records as below
1 - 07/23/2013 - 07/24/2013
2 - 07/15/2013 - 07/25/2015
3 - 03/10/2013 - 03/10/2014
Now i want to search these records by date range and want to see if some life exists in that range; e.g. i want to find the lives between 08/01/2013 - 01/01/2014
As expected result it should select the life#2 and life#3
How can this be done with MySQL query?
Any help is highly appreciated.
This query should do it:
SELECT
*
FROM
lifecycles
WHERE
str_to_date(life_start_date, '%m/%d/%Y') <= '2014-01-01'
AND str_to_date(life_end_date, '%m/%d/%Y') >= '2013-08-01';
Which basically means life hasn't started before the end of the range you are looking for, and life didn't end before the range start.
Since you keep dates in VARCHAR format, you need to use str_to_date function, which is bad since MySQL won't be able to utilize any possible indexes you have on start_date or end_date columns.
This might help you.
SELECT SUM( IF( '2014-01-02' BETWEEN from_date AND to_date, 1, 0 ) ) AS from_exist,
SUM( IF( '2014-02-12' BETWEEN from_date AND to_date, 1, 0 ) ) AS to_exist
FROM date_range
So based on the results you can check whether date is between existing date range or not.
So you want to exclude lifes that are ended BEFORE 08/01/2013 and the ones that are not started AFTER 01/01/2014. This should work:
SELECT *
FROM lifecycles as alive
WHERE NOT EXISTS (
SELECT 1
FROM lifecycles as dead
WHERE dead.life_id = alive.life_id
AND (str_to_date(life_start_date, '%m/%d/%Y') > '2014-01-01'
OR str_to_date(life_end_date, '%m/%d/%Y') < '2013-08-01'))
I have the following data in my table. BTW ... this is a DD/MM/YYYY format:
Date
18/09/2012
17/09/2012
13/09/2012
11/09/2012
10/09/2012
09/09/2012
25/08/2012
24/08/2012
The result what I want are:
Date
18/09/2012
13/09/2012
11/09/2012
09/09/2012
25/08/2012
The rule:
It starts from the latest date (18/09/2012) and check the next one down (17/09/2012). If there is a date then removed that from the list because it requires to have 1 day apart. Then goes to 13/09/2012 and then check 12/09/2012 and didn't find and then move to next one so on and so on. Basically you can't have date close each other (min 1 day apart).
Now I can do this on cursor if it's on TSQL however since I'm working on MySQL, is there any such thing in MySQL? Or perhaps any sub-queries approach that can solve this query?
I'm appreciated your feedback.
Try this solution -
SELECT date FROM (
SELECT
date, #d := IF(#d IS NULL OR DATEDIFF(#d, date) > 1, date, #d) start_date
FROM
dates,
(SELECT #d:=null) t
ORDER BY
date DESC
) t
WHERE start_date = date
The subquery finds out start days (18, 13, 11...), then WHERE condition filters records. Try to run the subquery to understand how it works -
SELECT
date, #d := IF(#d IS NULL OR DATEDIFF(#d, date) > 1, date, #d) start_date
FROM
dates,
(SELECT #d:=null) t
ORDER BY
date DESC
SELECT
"MyTable1"."Date"
FROM
"MyTable" AS "MyTable1"
LEFT JOIN "MyTable" AS "MyTable2" ON
ADDDATE("MyTable1"."Date", INTERVAL 1 DAY) = "MyTable2"."Date"
WHERE
"MyTable2"."Date" IS NULL
ORDER BY
"MyTable1"."Date" DESC
As long as I know about mysql query will be quit tricky and buggy if some how you manage to write the one. I suggest go for cursor, here is the syntax of the cursor,
here is the syntax of the cursor
I am designing a report where I need to display all days in a month(ex: 1- 30 for month of april) and display information associated to it. But my report displays only dates that have data associated to it and ignores all other days. Is there any way to display all the days in the report irrespective of the data?
As Wil hints at in his comment, it is easiest to generate these days in your dataset/query. This is a perfect place to use a CTE (Common Table Expression). These generate a table that you can use in your query, and they support recursion, so it is easy to generate a table with as many days as you want.
A Google of "cte days in a month" gave me this link as one of the top results: http://smehrozalam.wordpress.com/2009/06/09/t-sql-using-common-table-expressions-cte-to-generate-sequences/
Please leave a comment if this doesn't make sense.
OK, here's a sample using a CTE with days of the current month.
WITH DaysOfMonthCTE
AS ( SELECT DATEADD(month, DATEDIFF(MONTH, 0, GETDATE()), 0) AS StartDate ,
DATEADD(month, DATEDIFF(MONTH, 0, GETDATE()), 1) AS EndDate
UNION ALL
SELECT DATEADD(day, 1, StartDate) ,
DATEADD(DAY, 1, EndDate)
FROM DaysOfMonthCTE
WHERE EndDate < DATEADD(month,
DATEDIFF(MONTH, 0, GETDATE()) + 1, 0)
)
SELECT Student.Name ,
DaysOfMonthCTE.StartDate ,
AVG(Scores.TestScore) AS AverageScore
FROM myStudentsTable AS student
INNER JOIN myScoresTable scores
ON student.ID = scores.StudentID
RIGHT OUTER JOIN DaysOfMonthCTE
ON scores.TestDate >= DaysOfMonthCTE.StartDate
AND scores.TestDate < DaysOfMonthCTE.EndDate
GROUP BY Student.NAME ,
DaysOfMonthCTE.StartDate
Answer by #JamieF is spot-on but I was looking for a more general purpose date generator where I can get a list of dates between a start date and end date. The OP asked for all the days in a month, so this answer is not really for the OP. I added here in the event someone (like me) is looking for how to generate sequential data in a range on the fly in SSRS.
I merely added the following as the basis for my query in a dataset, and used two parameters - #StartDate and #EndDate.
;with DateGenerator
as (
select
CONVERT(datetime, #StartDate) GeneratedDate
union all select
DATEADD(day, 1, GeneratedDate)
from
DateGenerator
where GeneratedDate < #EndDate
)
select
GeneratedDate
from
DateGenerator
option (maxrecursion 32767)
Beware - limit is range of 32767 days. If the range is larger than that this solution will implode.
To use, merely join your query to the pseudotable DateGenerator (line above 'option' statement)