I have the following, which is used to setup variables throughout a project:   %macro dateset; %global cmonth pmonth cyear pyear cmonth_name; %let sysmonth = %sysfunc(month("&sysdate"d)); %let sysyear = %sysfunc(year("&sysdate"d)); %if &sysmonth = 1 %then %do; %let cmonth = 12; %let cmonth_name = "December"; %let pmonth = 11; %let cyear = %sysevalf(&sysyear - 1); %let pyear = &cyear; %end; %if &sysmonth = 2 %then %do; %let cmonth = 1; %let cmonth_name = "January"; %let pmonth = 12; %let cyear = &sysyear; %let pyear = %sysevalf(&cyear - 1); %end; %if &sysmonth >= 3 %then %do; %let cmonth = %sysevalf(&sysmonth - 1); %let cmonth_name = ????? %let pmonth = %sysevalf(&sysmonth - 2); %let cyear = &sysyear; %let pyear = &cyear; %end; %mend; %dateset; %put &cmonth &cyear &cmonth_name &pmonth &pyear; I am stuck with converting the month number in:     %let cmonth = %sysevalf(&sysmonth - 1); %let cmonth_name = ?????   Probably easy for some here....but I am stuck.  Thanks ... View more

I am trying to fix up a SAS EG project that I have "inherited"   There is a code node that is necessary at the start to assign dates.  The user will open the code node and enter the number of days back in time they wish to use.  So the following is used:   %let numdays=8;   The code that is there now identifies the start and end date for variable selection later (based on date).  So two variables are created; a start date and end date. The current code, which I believe is wrong is:   %let startdt=datetime()-(&numdays*(60*60*24)) + (170)*60; %let enddt=(today());     First off, I believe the last line should be:   %let enddt=%sysfunc(today());   So this leaves the startdt to fix. I am not sure at all what the code should be.  All I need is for the start date to be 8 days ago (or whatever the number) from today. In any case, you will notice that there are no sysevalf or sysfunc in the original code, so it is wrong anyway.  Thanks. I have no idea why the 170*60 is there and cannot ask the person who wrote it why it is in place.           ... View more

We have flight data that looks like:   SPECIALTY     FLT_BIZ_UID    FIR_NAME     DISTANCE 1             1234567        MONCTON      25 2             1234567        MONCTON      40 3             1234567        MONCTON      60 4             1234567        TORONTO      25 5             1234567        TORONTO      40 6             1234567        TORONTO      80 8             1234567        VANCOUVER    25 7             1234567        VANCOUVER    40 6             1234567        VANCOUVER    80   I would like to be able to select the first entry for each flight (flt_biz_uid) as it enters a different FIR (flight information region).   So the output would be:   FLT_BIZ_UID    FIR_NAME     1234567        MONCTON      1234567        TORONTO      1234567        VANCOUVER   All this is doing is getting a track of a flight as it moves into different FIR regions (there can be many flights...many flt_biz_uid).  I had tried using .first, but could not get the syntax to work.   As a part two, is it possible to do the same thing but add a total distance column for each FIR?  For example:   FLT_BIZ_UID    FIR_NAME   DISTANCE  1234567        MONCTON    125 1234567        TORONTO    145 1234567        VANCOUVER  145   thanks in advance ... View more

When aircraft fly in our airspace they are given a unique numeric ID called FLT_BIZ_UID. As messages are sent between the aircraft and ATC, they are stored (items stored include, the message, aircraft type and a bunch of other things). Every time there is a message by that specific flight a new FLT_BIZ_UID record is made (with the same numeric number of course as the flight is continuing on).   For example:   FLT_BIZ_UID     15 variables across    EDMONTON_DATALINK    GANDER_DATALINK     GANDER_VOICE   123456                                                   1                                           0                                       0                            123456                                                   1                                           0                                       0 333333                                                   1                                           0                                       0 333333                                                   1                                           0                                       0 333333                                                   1                                           0                                       1   So as you can see, there are multiple records for FLT_BIZ_UID based on the number of times the aircraft messages ATC   I would like to be able to generate a subset of this data.   I want all FLT_BIZ_UID observations that are all EDMONTON_DATALINK=1 GANDER_DATALINK=0 and GANDER_VOICE=0   So above, all 123456 records would be kept, while none of 333333 would be kept because the last observation of the group containing 333333 in the FLT_BIZ_UID has a GANDER_VOICE=1   Not sure how to approach this   thanks ... View more

This is probably easy, but maybe not. I have data as:   origin    destination   distance_km CYYC   CYUL           100 EGGL   EDDF            200   They airport combinations (origin and destination). Many occur and can repeat.     How could I get output of unique airport combinations with the average km?   For example, the output would show for every unique origin/destination pair   origin    destination   avg_distance_km CYYC   CYUL          101.554 EGGL   EDDF           202.678     thanks ... View more

I have 41 variables I need to calculate a growth rate for. I have them defined in a macro loop but I don't think it can be done in a datastep.  I have:   data TAFOASAS.TAM_TWRS_FREQ_ALL_G; set EGTASK.TAM_TWRS_FREQ_ALL; CYYZ_FREQ_G = CYYZ_FREQ / lag12(CYYZ_FREQ) -1; format CYYZ_FREQ_G PERCENT10.2; drop cyyz_freq; run;   The only thing that changes is the airport code (CYYZ in this case). Can a list be passed to create the growth rate in a loop for each airport?   thx. ... View more