I have a dataset with daily data. I need to create a variable which brings the number of the week in each month (1 to (4 or 5)).
I have reached that with the following formula:
WK_NUM = intck('week',intnx('month',DATE,0),DATE)+1;
It works fine, but frequently the first and the last weeks of the month aren't full weeks, so the line only brings partial weeks. I need to improve it to bring remaining days from previous or next month in order to fill each week with 7 days.
Any ideas?
Hope this creates the expected result:
/* initial data set with dates */
data a;
format date date9.;
do date="15apr2017"d to "15jun2017"d;
output;
end;
run;
/* adding week number */
data a;
set a;
wk_num = week(date,'u');
run;
/* selecting May and modifying wk_num */
proc sql noprint;
CREATE TABLE b AS
SELECT date
,wk_num - ((SELECT min(wk_num) FROM a WHERE date="01may2017"d)-1)
FROM a
WHERE wk_num between (SELECT min(wk_num) FROM a WHERE date="01may2017"d) AND (SELECT max(wk_num) FROM a WHERE date="31may2017"d)
;
quit;
Option for week function:
u (default): specifies the number-of-the-week within the year. Sunday is considered the first day of the week. The number-of-the-week value is represented as a decimal number in the range 0-53. Week 53 has no special meaning. The value of week('31dec2006'd, 'u') is 53.
v: specifies the number-of-the-week whose value is represented as a decimal number in the range 1-53. Monday is considered the first day of the week and week 1 of the year is the week that includes both January 4th and the first Thursday of the year. If the first Monday of January is the 2nd, 3rd, or 4th, the preceding days are part of the last week of the preceding year.
w: specifies the number-of-the-week within the year. Monday is considered the first day of the week. The number-of-the-week value is represented as a decimal number in the range 0-53. Week 53 has no special meaning. The value of week('31dec2006'd, 'w') is 53.
Related
Currently have rows aggregated by week number.
SELECT to_char(date, 'IYYY-MM-IW') AS week, from TABLE GROUP BY week
The results will show the form "2021-07-29". Is it possible to change the week number such that it is the number week of the month (instead of year).
For example, instead of "2021-07-29", we convert to "2021-07-04" since the 29th week of the year is actually the 4th week of the month.
Quote from the manual
W week of month (1–5) (the first week starts on the first day of the month)
So you can use:
to_char(date, 'YYYY-MM-W')
For e.g 2021-10-18 this yields 2021-10-3 (third week in October)
I'm used to do the following syntax when analysing weekly data:
select week(creation_date)::date as week,
count(*) as n
from table_1
where creation_date > current_date - 30
group by 1
However, by doing this I will get just part of the first week.
Is there any smart way to alway get a whole week in the beginning?
Like get the first day of the week I would get half of.
First off you need to define what you mean by "week". This is more difficult than it appears. While humans have an intuitive since of a week, computers are just not that smart. There are 2 common conventions: the ISO-8601 Standard and, for lack of a better term, Traditional. ISO-8601 defines a week as always beginning on Monday and always containing 7 days. Traditional weeks begin on Sunday (usually) but may have weeks with less than 7 days. This results from having the 1st week of the year beginning on 1-Jan regardless of day of week. Thus the 1st and/or last weeks may have less than 7 days. ISO-8601 throws it own curve into the mix: the 1st week of the year begins on the week containing 4-Jan. Thus the last days of Dec may be in week 1 of the next year and the first days Jan may be in week 52/53 of the prior year.
All the below assume the ISO-8061.
Secondly there is no week function in Postgres. In you need extract function. So for this particular case:
select extract(week from creation_date)::integer as week, ...
Finally, your predicate (current_date - 30) ensures you will unusually not begin on the 1st of the week. To get the correct date take that result back 1 week, then go forward to the next Monday.
with days_to_monday (day_adj) as
( values ('{7,6,5,4,3,2,1}'::int[]) )
select current_date - 30
, current_date - 30 - 7 + day_adj[extract (isodow from current_date - 30 )]
from table_1 cross join days_to_monday;
The CTE establishes an array which for a given day of the week contains the number of days need to the next Monday. That main query extracts the day of week of current date and uses that to index the array. The corresponding value is added to get the proper date.
Putting that together with your original query to arrive at:
with next_week (monday) as
( values (current_date - 30 - 7
+ ('{7,6,5,4,3,2,1}'::int[])[extract (isodow from current_date - 30 )])
)
select extract(week from creation_date) as week,
count(*) as n
from table_1
where creation_date >= (select monday from next_week)
group by 1
order by 1;
For full example see fiddle.
I need to convert Week to start with Sunday (range 1-53, with a Sunday in this year)
Ex. DATE_PART(w,'2020-01-05') (Sunday)
This should return 2 instead of 1.
Thanks in advance.
If you wish to obtain the Week Number where a week starts on Monday instead of Sunday:
Subtract one day to the date being convert to the Week Number
So, use:
DATE_PART(w, date_field - INTERVAL '1 DAY')
This returns the Week Number of the 'previous day'.
Unfortunately, it won't work for the first day of the year, which will appear to be the last day of the previous year. Depending on how your organization treats the first/last weeks of the year, this might be okay.
If that is not satisfactory, then you could write your own User-Defined Function (UDF) and code it to return the values you desire.
Creating a function that takes a start and end date and counts how many Sundays between those dates fell on the 1st of the month on kdb+, how would I do this?
The function needs to show how many times this has happened since 1950
Let's define a function which returns a weekday of its argument (of type date) first.
The underlying value of a date is the count of days from 1/1/2000 and we know that 1/1/2000 was Saturday. The next day was obviously Sunday, then Monday etc. and every 7th, 14th, 21st, etc. day after and before Jan 1, 2000 was Saturday too. So if we take a date modulo 7 we'll get a weekday number where 0 is Saturday, 1 is Sunday, etc. which leads us to the following definition.
weekday:{ `sat`sun`mon`tue`wed`thu`fri x mod 7 }
Now we can create a function that answers the original question:
sundaysThe1st:{[start;end]sum `sun=weekday dates where 1=`dd$dates:start+til 1+end-start }
start+til 1+end-start generates a list of dates between start and end, dates where 1=`dd$dates returns only the first days of the months and `sun=weekday dates returns 1b if the 1st day of the month is Sunday and 0b otherwise. sum is effectively the number of 1's which is exactly what we need.
Hope this helps.
I'm trying to separate weeks from timestamp per quarter so it should be between 1-13 week per quarter so I used function week() but it takes between 1-52 week as whole year so I made it to be divided by function of quarter like below
select Week (EVENTTIMESTAMP) / QUARTER (EVENTTIMESTAMP) from KAP
The thing here that results aren't accurate; for example it shows:
time stamp 2014-07-06 12:13:03.018
week number 9
which isn't correct because July is first month in Q3 and it's in the 6 days so it should be 1 week from Q3 not 9.
Any suggestion where it go wrong?
You want something like WEEK modulo 13 to get week number within a quarter. You will have to tinker with 'modulo 13 yields 0..12' by adding or subtracting one at appropriate points.
Some minimal Google searching using 'ibm db2 sql modulo' yields DB2 MOD function:
The MOD function divides the first argument by the second argument and returns the remainder.
Hence MOD(WEEK(...), 13), except you probably need MOD(WEEK(...)-1, 13) + 1, as intimated already.
You may need to watch for what the WEEK() function does at year ends:
The WEEK function returns an integer in the range of 1 to 54 that represents the week of the year. The week starts with Sunday, and January 1 is always in the first week.
I'm curious about how they can come up with week 54. I suppose it requires 1st January to be a Saturday (so 2nd January is the start of week 2) of a leap year, as in 2000 and 2028. Note that week 53 and (occasionally) week 54 will show up as weeks 1 and 2 of Q5 unless you do something. Also, Saturday 2000-03-25 would be the end of Q1 and Sunday 2000-03-26 would be the start of Q2 under the regime imposed by the WEEK() function and a simple MOD(WEEK(...), 13) calculation. You're likely to have to tune this to meet your real requirements.
There's also the WEEK_ISO() function:
The WEEK_ISO function returns an integer in the range of 1 to 53 that represents the week of the year. The week starts with Monday and includes seven days. Week 1 is the first week of the year that contains a Thursday, which is equivalent to the first week that contains January 4.
Note that under the ISO scheme, the 3rd of January can be in week 52 or 53 of the previous year, and the 29th of December can be in week 1 of the next year. Curiously, there doesn't seem to be a YEAR_ISO() function to resolve such ambiguities.
In a data warehouse, the proper solution to this is to create a time dimension that contains static mappings for days/weeks/months/quarters/years. This provides the ability to define these based on your business' fiscal calendar (if it is not following on the calendar year).
See: http://www.kimballgroup.com/1997/07/10/its-time-for-time/ for more information.