@Alice_SAS wrote: Yes, that situation exists. If the two separate date ranges don't overlap, I prefer not to merge them and leave them as they are.  Please post data containing this case and post the data in usable form. ... View more

Following Maxim 14, i would use "find in files" in notepad++ ...   ... View more

if you don't need the result sorted, using a hash object provides you with dataset want in one step: data want; set main; if _n_ = 1 then do; if 0 then set addon; declare hash h(dataset: 'addon'); h.defineKey('reprisk_id'); h.defineData('xxx'); h.defineDone(); end; if h.find() = 0; run; or change the merging step, so that an in-variable is defined for the addon dataset, too: data want_ds; merge main(in = inMain) addon(in = inAddon); by reprisk_id; if inMain and inAddon; run; I was not able to sort the data in the order of your want dataset. ... View more

Interesting. Is the error shown if the following code is executed in a fresh session? data stocks; set sashelp.stocks; emphasis = 2; if stock eq 'Microsoft' then emphasis=1; run; title 'Microsoft Compared to IBM and Intel'; proc sgplot data=stocks (where=(date >= "01jan2003"d)); series x=date y=close / group=stock grouplc=emphasis; run; title; ... View more

Assuming that "and BMI_group ..." belongs to the where statement. proc print data=homework_1; where sex='F' AND BMI_group='overweight'; run; Attaching a format to a variable does not change the type of the variable, only the way its values are displayed changes. So BMI_group is still numeric and the comparison with "overweight" causes the error. ... View more

Create an new timestamp using the datepart of end_dt_time to compare it with the start and end: Want = (beg_dt_time <= dhms(datepart(end_dt_time), 1, 0, 0) <= end_dt_time); ... View more

Could be solved by a data step, too, but i don't think, that it is better. Just to clarify: you don want one obs for each source/target combination with the earliest start date and the latest end date? ... View more

I am happy to help, even with homework task, but i except that some efforts are shown. In this case: post the data as working data step, so that we have something to test our code post your ideas ...   ... View more

I would have a closer look at the step creating the first table, maybe it can be improved to avoid fixing it. ... View more

No the syntax is not correct. You have to compare each variable with the list of values: where var1 in (1, 2, 3, 4) or var2 in (1, 2, 3, 4) .... ... View more

Maybe you should start following Maxim 1 This will save you time. ... View more

Step 1: Fix the data type of the variable date, changing it to a numeric variable with a date format attached. Untested: data work.date_fixed; set work.test(rename=(date = str_date)); date = input(str_date, mmddyy10.); format date date9.; drop str_date; run; Step 2: Use a retained variable to store the date variable when phase = "closed", a subsetting-if can be used to keep only obs with date > retained_date. If you have multiple ids in the data, i am sure your need retained_date set to missing if the id changes. ... View more

Contact the mssql admin, afaik granting read access to a view is not sufficient, read access to the used tables is necessary, too. ... View more

Can you post the array statement, at least? If you can't, it will hardly possible to help you. ... View more