match event windows (in month) using sas sql - sql

I have two files, one with event date, the other one is the security prices I need to match the month with event month 0, -1 (previous month), and +1 (the month after event month). So e.g., one observation has event date 1/1/2010, I have the daily bond prices from 1/1/2009 to 1/20/2010, but I need to match January,2010 to December 2009 prices and February 2010. I have the following codes to do that,
PROC SQL;
CREATE TABLE test AS
select *
FROM bondprice as a, sdc as b
where b.Participant_CUSIP=a.cusip
& INTCK('MONTH',b.Alliance_Date,a.time)>=-1
& INTCK('MONTH',b.Alliance_Date,a.time)<=1 ;
QUIT;
DATA test1;
SET test;
BY deal_number CUSIP code time;
IF first.code THEN price1=price;
IF first.code THEN day1=time;
IF last.code THEN price0=price;
IF last.code THEN day0=time;
RUN;
clearly this observation should be only matched in 1-month Jan 2010 (0,0) and 2-month Dec 2009 to Jan 2010 (-1,0), because it does not have 3-month Feb 2010's prices. However, if using my codes above, it will yield as a valid obs who has 3-month price since it fell into the range of >=-1 and <=1.
So the first part of code is not accurate (the >= and <= range): it yields any prices that falls into December 2009 to Feb 2010, this is wrong because of two situations. First, if the bond prices stops in between Dec 2009 and Jan 2010, it will be counting as a valid 2-month and 3-month observations. Second, if the bond prices begins in between Jan 2010 and Feb 2010, it will be counting as a valid 3-month observation. Both of these two situation will yield invalid observations, since the first situation do not have Feb 2010 prices but still counts as 2-month and 3-month, and the second situation do not have either Dec 2009 or Jan 2010 prices, but still counts as 3-month.
Will someone help me with this? How can I modify the code so that the 3-month window and 2-month window contains the relative months?
I hope I describe the problem clearly. Please let me know if anything is unclear.
I've uploaded an example to this problem here.
https://www.dropbox.com/s/hx9ahst7nts4k0q/ex1.xlsx?dl=0

Related

SQL - check if an order date occurs after the second Saturday in July

I am querying against a table of 4 yrs of order transactions (pk = order number) and I'm looking to tag each record with particular date flags based on the order date - e.g., calendar year, calendar month, fiscal year, etc. There are date attributes that are specific to our business (e.g., not easily solved by a datepart function) that I'm having trouble with.
I was able to add "School Year" (for us that runs Aug 1 - July 31) using a case statement:
case
when datepart(month, oline.order_date_ready) between 8 and 12 then datepart(year, oline.order_date_ready)
else (datepart(year, oline.order_date_ready)-1)
end as school_yr
So for 1/19/2017, the above would return "2016", because to us the 2016 school year runs from Aug 1 2016 to July 31 2017.
But now I'm having trouble repeating the same kind of case statement for something called "Rollover Year". All of our order history tables are reset/"rolled over" on the 2nd Saturday in July every calendar year, so for example the most recent rollover date was Saturday July 9th 2016. Click to view - rollover year date ranges
My above case statement doesn't apply anymore because I can't just add "datepart(month, oline.order_date_ready) = 7" - I don't need the whole month of July, I just need all the orders occurring after the 2nd Saturday in that July. So in this example, I need everything occurring from Sat July 9 2016 to today to be flagged as rollover_date = 2016.
Is there a flexible way to do this without hard coding previous/future rollover dates into another table? That's the only way I can think to solve it currently, but I'm sure there must be a better way.
Thanks!
If you ask for the day-of-the-week of July 1st, then from there it's simple arithmetic, right? This query gives results matching your image:
SELECT y,
CONCAT(y, '-07-01')::timestamp +
CONCAT(6 - EXTRACT(DOW FROM CONCAT(y, '-07-01')::timestamp) + 7, ' days')::interval
FROM generate_series(2013, 2020) s(y)
ORDER BY y DESC
;
So given any date d from year y, if it comes before the 2nd Saturday of July, give it fiscal year y - 1. Otherwise give it fiscal year (school year?) y.

T-SQL: August is the beginning of next years financial year, code to dynamically to change the batch ids

I have a column called batch_id with a list of dates - 2016080184 ie date 2016 08 01 84(84, I believe a time part).
I need to update the batch_id (varchar(25)) to change to 2017010184, based on another column voucher_date (datetime) = 2016-08-01 00:00:00.000
So if the voucher date was 2016-08-02 00:00:00.000, then the batch_id needs to change from 2016080278 to 2017010278 (78 at the end here doesn't matter)
August is the first month for the financial year, so August would effectively become January, September would become February etc.. and the year from Aug needs to indicate the following year ie this year is 2016 therefore batch_id should start with 2017.
Next August batch_id should indicate 2018 etc..
The file I receive is always a day behind to make things more complicated.
I am a little confused about your requirement for the year to change, but this should give you all you need to get started:
declare #BatchID nvarchar(10) = '2016080184'
select convert(nvarchar(8),dateadd(month,-7,cast(left(#BatchID,8) as date)),112) + right(#BatchID,2) as NewBatchID_SameYear
,convert(nvarchar(8),dateadd(month,5,cast(left(#BatchID,8) as date)),112) + right(#BatchID,2) as NewBatchID_NextYear
You can apply the above date changes with a CASE statement on the voucher_date column as required.

Powerpivot: year to date vs year to date previous year

I am using excel 2013 powerpivot which I have linked to a sql query. In this query I have all sales data from previous years and it updates itself with the current sales. I want to make a pivottable showing year to date sales from this year compared to the same period last year. So for example from Jan.1st 2015 until July 10th of 2015 compared to Jan.1st 2014 until July 10th 2014.
I linked my sales data table with a calendar table. But whatever I try, parallelperiod, sameperiodlastyear, totalytd, it always shows me the correct data for this year, but the full year sales of last year. Can anyone recommend me what to try?
thanks,
Frank
The TOTALYTD (and other time intelligence functions) will look for the highest date in whichever context you give it, so when you try to tell it to go back 12 months from now it thinks "OK, I'm in 2014" I'll get ALL the data from 2014 and calculate a TOTALYTD. So you have to ignore the built-in functions and build your own:
=CALCULATE(sum(Table1[sales]),DATESBETWEEN(DateDim[Date], FIRSTDATE(DATEADD(DateDim[Date],-12,MONTH)), LASTDATE(DATEADD(Table1[Date],-12,MONTH))))
So long as you have a slicer or a field on your table that picks up the year, then that should work.
Edit: Having tested with a data sample, again DAX is trying to be too clever and because I have told it to go back 12 MONTHS, it assumes I want ALL data from the month context which is including everything from July last year.
Going back 365 days fixes this (as long as it's not a problem for leap years).
=CALCULATE(sum(Table1[sales]),DATESBETWEEN(DateDim[Date], FIRSTDATE(DATEADD(DateDim[Date],-12,MONTH)), LASTDATE(DATEADD(Table1[Date],-365,DAY))))
I have tried your way, but it does not yet give me the correct answer. Let me show you what I have right now.
For current year sale I have the following Dax formula
=TOTALYTD(sum(Omzetgegevens[NettoOmzet]); Kalender[Calender date])
For the previous year sale I have:
=CALCULATE((SUM(Omzetgegevens[NettoOmzet])); SAMEPERIODLASTYEAR(DATESYTD(Kalender[Calender date])))
To test your solution I have called "test ytd":
=CALCULATE(SUM(Omzetgegevens[NettoOmzet]); DATESBETWEEN(Kalender[Calender date]; FIRSTDATE(DATEADD(Kalender[Calender date];-12; MONTH)); LASTDATE(DATEADD(Kalender[Calender date]; -365; DAY))))
If I run the pivottable now, the result I get is:
2015
current year ytd last year ytd test ytd
januari 28.912 34.487 34.487
februari 50.301 66.003 31.516
maart 73.362 92.647 26.644
april 99.561 117.853 25.205
mei 128.021 149.108 31.255
juni 149.706 174.076 24.968
juli 158.297 205.438 31.362
augustus 158.297 231.693 26.255
september 158.297 254.534 22.841
oktober 158.297 282.484 27.951
november 158.297 303.808 21.324
december 158.297 313.208 9.400
Total 158.297 313.208 313.208
What I would like to see is the following. As this report was run on the 14th of july 2015. I want to see all sales for 2015 until this date and all the sales for 2014 until july 14th 2014.
If it is not possible to see the months, I am also fine with only a total number for current ytd and last year ytd
2015
current year ytd last year ytd
januari 28.912 34.487
februari 50.301 66.003
maart 73.362 92.647
april 99.561 117.853
mei 128.021 149.108
juni 149.706 174.076
juli 158.297 175.312(so not full month of july in 2014)
Total 158.297 175.312
Recently had a similar issue. Current context confuse a lot and did not return the result we think should be returned. Play with DATEADD to go back and forth in combination with any of the date/time function. You will get the result and the context will also be clear.

MS Access grouping query spanning start and end dates

I would like to get a running tally of how many widgets were/are rented at any one time, by month, by year. Data is held in an MS Access 2003 db;
Table name: rent_table
Fields:
rentid
startdate
enddate
rentfee
rentcost
bookingfee
Something like; Count number of rentid's that fall between month/year, then group them?
e.g. if a widget was rented from 5th Jan 2014 to 8th April 2014 it would appear as a count in Jan, Feb, Mar and April tally's.
Many thanks.
EDIT
More details (sorry);
Access db is fronted by classic ASP.
If possible I don't want to create any new tables.
No input is required in order to run the report.
There are around 350-400 widgets that could be rented at any one time.
Each widget is rented exclusively.
Report output example;
Month | Year | NumRented
Jan 2014 86
Feb 2014 113
...
Can a query pick up dates within dates? So literally do a count of the table where date >Dec 31st 2013 AND <1st Feb 2014 (to grab a count for all of January 2014) and would that include the example of the rent starting on the 5th Jan? So I could just do twelve counts for each year?
create a calendar table, e.g.
table = cal_yyyymm with one column dt_yyyymm as numeric field
populate the table with ... say 5 or 10 years of data
201401 201402 201403 ... 60 or 120 rows, a small table
make a sql
Select
dt_yyyymm,
count(*) as cnt
From cal_yyyymm
Left Join rent_table
On format(startdate,"yyyymm") >= dt_yyyymm
And dt_yyyymm >= format(enddate,"yyyymm")
think about the complications in the data -- --
widget was rented from 5th Jan 2014 to 8th Jan 2014
and again rented from 11th Jan 2014 to 21st Jan 2014
does this count at 1 or 2 in the month?
if it is 1, then the sql gets more complicated because
the rent_table first needs to have its dates converted
to yyyymm format, and second needs to be de-duped on rentid,
and third then joined to cal_ On the dates...

Lookup a value from a range of dates in excel

Please help.
We have two tables. A list of accounts and the other is a change log. I need to add a new column in table 1 where the value is the amount in table 2 for the correct account and correct validity period.
ex.
table 1:
account# beginning period ending period
1 January 1, 2012 January 31, 2012
2 January 12, 2012 February 12, 2012
table 2:
account # amount valid period beg valid period end
1 10 january 1, 2009 december 5, 2010
1 20 december 6, 2010 june 1, 2011
1 30 june 2, 2011 december 1, 2012
2 13 january 15, 2011 december 15, 2011
2 20 december 16, 2011 february 20, 2012
Thanks.
Although it is a bit complex requirement it could be done with built-in functions (although it can look a bit obscure :-) ). Specifically I mean function SUMIFS.
It has several parameters.
The first one is an area with values to be summed. It is B8:B12 in this example.
The second is an area whith values to be checked with some condition. It is A8:A12.
The third is a criterion to be applied for area from second parameter. It is (inter alia) account #.
So the formula says: sum all values from rows in B8:B12 where account # (A8:A12) is equal to desired account # (e.g. A3).
Ok, it is not all, you need specify the time range. It would be a bit clunky because you must check if two time period are overlapping (it would be easier to check if one date is in specified period).
But it could be done because SUMIFS can take another pairs of criteria range and criterion. But it cannot be used for OR condition, so you had to combine more SUMIFS.
Nice article about overlapping ranges is e.g. http://chandoo.org/wp/2010/06/01/date-overlap-formulas/
BTW: you have to format cells in B2:C3 and C8:D12 as a date to be able to compare them.