cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
JoMa
‎03-14-2017
  • 4 Posts
  • 0 Likes Given
  • 0 Solutions
  • 0 Likes Received

Re: Conditional lag

by Fluorite | Level 6 in SAS Programming
‎02-14-2016 03:03 PM
‎02-14-2016 03:03 PM
Thanks a lot!!!!! ... View more

Re: Conditional lag

by Fluorite | Level 6 in SAS Programming
‎02-14-2016 11:37 AM
‎02-14-2016 11:37 AM
oh my! it was as simple as that. Thanks for your help ... View more

Re: Conditional lag

by Fluorite | Level 6 in SAS Programming
‎02-14-2016 08:20 AM
‎02-14-2016 08:20 AM
I'll  try to explain it by userid=444.    Userid=444 has the following values (from the above table): Row UserID Category Day Values PosVal(d-1) PosVal(d-2) 1 444 1 1 4 0 0 2 444 1 2 42 4 0 3 444 2 1 4 0 0 4 444 2 2 2 4 0 5 444 2 3 7 2 4   Values is a column and the content of it should be lagged.   PosVal(d-1) is the content of “ Values”  from the day before (d-1) of userid  X in category X If there is no row of UserIDX on the previous day then PosVal(d-1)=0 Example row 1: PosVal(d-1) = content of “ Values”   of userid444 on day=0 in category 1 -> day 0 does not exist, hence=0 Example row2= PostVal(d-1) content of “ Values”    of userid444 on day=1 (day before day 2) in category 1 is 4;   PosVal(d-2) Is the content of “ Values”   from two days before (d-2) of userid 444 in category If there is no row of UserIDx two days before then PosVal(d-2)=0 Example row 1: PosVal(d-2) = content of “ Values”    of userid444 on day= -1 in category 1 -> day -1 does not exist, hence=0 Example row 5: PosVal(d-2) = content of “ Values”    of userid444 on day= 1 (two days before day3) in category1 is 4   I am pretty much stuck in this misery...and appreciate any further help.... ... View more

Conditional lag

by Fluorite | Level 6 in SAS Programming
‎02-14-2016 06:54 AM
‎02-14-2016 06:54 AM
Hi there, I would need your help in restructuring a database based on a time series.   I have a database similar to this one (row-values): Row UserID Category Day Values PosVal(d-1) PosVal(d-2) PosVal(d-3) 1 970 1 1 3 0 0 0 2 444 1 1 4 0 0 0 3 333 1 1 5 0 0 0 4 234 1 1 1 0 0 0 5 970 1 2 2 0 0 0 6 555 1 2 3 0 0 0 7 444 1 2 42 4 0 0 8 456 1 2 7 0 0 0 9 970 1 3 5 2 0 0 10 333 1 3 31 0 5 0 11 123 1 3 4 0 0 0 12 321 1 3 63 0 0 0 13 6557 1 3 2 0 0 0 14 970 1 4 4 5 2 0 15 234 1 4 5 0 0 1 16 213 1 4 3 0 0 0 17 456 1 4 2 0 7 0 18 233 1 4 1 0 0 0 19 444 2 1 4 0 0 0 20 333 2 1 63 0 0 0 21 444 2 2 2 4 0 0 22 654 2 2 4 0 0 0 23 444 2 3 7 2 4 0   And PosVal(d-1), PosVal(d-2) and PosVal(d-3) are the variables that I need to create.     Basically these are conditional lags of the “values” depending on the userid, category and day where d-1 stands for the first lag, d-2 2 lags etc.  I believe that would require loops but I am very much unable to create these lags… I’d very much appreciate your help!   Thanks you   Jo ... View more
Contact Me
Online Status
Offline
Date Last Visited
‎03-14-2017 03:34 PM
Quick links