This question already has answers here:
Months between not returning correct value
(2 answers)
Closed 4 years ago.
I am using MONTHS_BETWEEN function to get the difference between 2 dates , but the result is not 100% correct i think this function mines 1 day .
i tried this SQL
SELECT MONTHS_BETWEEN(TO_DATE('20170630','YYYYMMDD'),
TO_DATE('20170501','YYYYMMDD') ) "Months"
FROM DUAL;
here in this case it should return 2 but the result is 1.93548387096774
So, Any help ?? or is there any way to add 1 more day to date??
Try
SELECT ROUND( MONTHS_BETWEEN(
TO_DATE('20170630','YYYYMMDD'),
TO_DATE('20170501','YYYYMMDD')))
"Months" FROM DUAL;
MONTHS_BETWEEN returns decimal result if days are different and they are not both specify the last day of the month. For more info see http://www.sqlines.com/oracle/functions/months_between
Months between will give you a decimal value with respect to a whole month. So from 1st Jan to 1st Feb, you will get 1. But from 1st Jan to 31st Jan you will not get exact 1. It will be 0.9.... Similarly from 31st Jan to 28th Feb (non leap year) you will get 1.
So you will get 2 from 1st May to 1st July OR from 30th April to 30th June,
From 1st May to 30th June, you will never get exact 2.
If you want exact 2, then either add 1 day to greater date or subtract 1 day from start_date.
DBFiddle Demo
Related
I am new to snowflake and my manager wants me to retrieve the data of the past month when it is 5th of the current month. For example if today is 5th April, then ask snowflake to retrieve the data of the past month i.e. from 1st March 2021 to 31st March 2021 and similar for all the other months.
The reason why he wants to update the last month data on 5th of every next month because that is the day when we get the data.
I tried to use the DATEADD function but it is more complicated than just using this function.
Thanks in advance!
PS: The data for every month has same date. for example: the date is like - April 20th will be stored in the database as "2021-4-01" - and same for April 25th date will be stored as "2021-4-01" .
The day doesn't change in the database, just the month and year.
as to the prior month window that can be done via DATE_TRUNC and DATEADD
select
current_date as cd
,date_trunc('month', cd) as end_range
,dateadd('month', -1, end_range) as start_range
;
gives:
CD END_RANGE START_RANGE
2021-04-21 2021-04-01 2021-03-01
the other half of the question only do it on the 5th, if you have a task run daily etc. can be solved via
,day(current_date) = 5 as is_the_fifth
or if in an inline way
iff(day(current_date) = 5, <do true stuff>, <do false stuff>)
I have a query which uses needs to know how many days passed since 1st of January in the current year.
Which means that if the query runs for example in:
2nd Jan 2017 than it should return 2 (as 2 days passed since 1st Jan
2017).
10th Feb 2016 than it should return 41 (as 41 days passed since 1st
Jan 2016).
basically it needs to take Current Year from Curent Date and count the days since 1/1/(Year).
i have the current year with: SELECT EXTRACT(year FROM CURRENT_DATE);
I created the 1st of Jan with:
select (SELECT EXTRACT(year FROM CURRENT_DATE)::text || '-01-01')::date
How do I get the difference from this date to Current_Date?
Basically this question can be Given two dates, how many days between them?
Something like age(timestamp '2016-01-01', timestamp '2016-06-15') isn't good because I need the result only in days. while age gives in years,months and days.
An easier approach may be to extract the day of year ("doy") field from the date:
db=> SELECT EXTRACT(DOY FROM CURRENT_DATE);
date_part
-----------
41
And if you need it as a number, you could just cast it:
db=> SELECT EXTRACT(DOY FROM CURRENT_DATE)::int;
date_part
-----------
41
Note: The result 41 was produced by running the query today, February 9th.
Given two dates, how many days between them
Just subtract one from the other.
In your case you could just round the current_date to the start of the year and subtract that from the current date:
select current_date - date_trunc('year', current_date)::date
The ::date cast is necessary to get the result as an integer, otherwise the result will be an interval.
Another solution is to use DATEDIFF
SELECT DATE_PART('day', now()::timestamp - '2016-01-01 00:00:00'::timestamp);
I am having a problem with making a Oracle Query that sums up everything that occurs during the whole last week and everything during this week. so lets say i have this table
DATA DATE
3 2-feb-15
4 3-feb-15
6 3-feb-15
7 27-jan-2015
5 27-jan-2015
4 25-jan-2015
so lets today is feb 5th and the query this week should be 3+4+6 assuming that sunday is feb 1st. The this week's data range should be (feb 2nd - feb 5th) minus sunday.
and the sum for last week should be 7+5+4 no matter what date today is as long as it still last week from Monday to saturday
Another case is, if todays date is feb 2nd, then the sum should be 3.
Please help me, i tried with SYSDATE - (Var) and it doesnt work because it doesnt position the date according to the calendar
thank you very very much
try this :
trunc(sysdate)
I hope its working if its not please give me more detail of your problem
Getting the month for the current date is, obviously, straight forward, but I'm needing to get the month name with a different end date.
I need to get the month name with the start date of the month being the first Thursday after the first Wednesday of the month and the end date of the month being the first Wednesday of the following month. It's for an accounting thing, so I'm not going to argue with the spec!
e.g. for 2014, January would run from 9th Jan - 5th Feb, February would run from 6th February - 5th March, March would run from 6th March - 2nd April.
I would suggest that you create a table with your 'accounting months' in it, having a start date, end date and month name columns.
You could then query this to find the row where your date is between the start and end dates and return the month name. Putting this into a scalar function would then allow it to be reusable and relatively easily updated for next years months as well.
I think, as per Paddy's answer, a lookup table is the simplest thing to do. Here's one way to generate the rows for it:
; With Numbers(n) as (
select 4 union all select 5
), Months as (
select CONVERT(date,'20010104') as StartDt,CONVERT(date,'20010207') as EndDt,
DATENAME(month,'20010103') as Month
union all
select DATEADD(week,n1.n,StartDt),DATEADD(week,n2.n,EndDt),
DATENAME(month,DATEADD(week,n1.n,StartDt))
from Months,Numbers n1,Numbers n2 --Old-skool join, just for once
where DATEPART(day,DATEADD(week,n1.n,StartDt)) between 2 and 8 and
DATEPART(day,DATEADD(week,n2.n,EndDt)) between 1 and 7 and
StartDt < '21000101'
)
select * from Months option (maxrecursion 0)
(CW since this is effectively just an extension to Paddy's answer but I don't want to edit their answer, nor is it suitable for a comment
I have a query with an IIF() expression in one column that I am using to identify if a date is previous to the current month to then amend it if so. So if I run the query on 19th March 2014 and the EffFrom date is before 1st March 2014, I would want that column entry to now appear as 1st March 2014.
I am using the below expression which is pretty much doing what I want, however I know it is not considering the year -- i.e. it is changing an entry of 1st Jan 2015 to be 1st March 2014.
EffFrom:
IIf(Month([Table.Efffrom])"Less than symbol"Month(Date()),Date()-Day(Date())+1,[Table.Efffrom])
Can someone correct the expression for me?
I interpreted "if I run the query on 19th March 2014 and the efffrom date is before 1st March 2014 I would want that column entry to now appear as 1st March 2014" to mean you want something like this from a query run today (Mar 19th 2014):
id Efffrom adjusted_date
1 1/1/2014 3/1/2014
2 3/1/2014 3/1/2014
3 3/31/2014 3/31/2014
4 1/1/2015 1/1/2015
If that is correct, your IIf expression can use DateSerial to check whether Efffrom is before the first of this month, and transform the older dates.
SELECT
y.id,
y.Efffrom,
IIf
(
y.Efffrom < DateSerial(Year(Date()), Month(Date()), 1),
DateSerial(Year(Date()), Month(Date()), 1),
y.Efffrom
) AS adjusted_date
FROM YourTable AS y;
Here's what i do to get the first day of the next month
EffFrom: DateAdd("m",1,CDate(Format([Table.Efffrom],"m/\1/yy")))