Computing The Variance Of Grouped Data
  Home FAQ Contact Sign in
 
 Advanced search
MATCHING GROUPS



more...
POPULAR GROUPS

more...
found 1491 articles for 1.761 sec
PS. If you want the mean or variance in price across months by group you could do: data temp; set data; avgprice=mean of ( price1,price2, ...price24) ; varprice=var of ( price1,price2, ...price24); run; proc summary NWAY data=temp; class group; var avgprice varprice; output out= xxx.temp mean=meanbygroup varbygroup; BILL WEST <wm_a_west@YAHOO.COM> wrote:     

Group: comp.softsys.sas · Group Profile · Search for Computing The Variance Of Grouped Data in comp.softsys.sas
Author: BILL WEST
Date: Jan 24, 2008 07:45

Randall, I think the problem you're facing is due to not using the option NWAY in your proc summary statement. Thus your results include the total mean and variance by month. So first thing I'd do is include the NWAY option to exclude the overall mean and variance. Also your example shows you are looking at variance across months rather than within groups. Proc summary NWAY with the class
Show full article (3.38Kb)
How do the other statistically minded people here feel about avoiding this whole issue by using permutation tests? Doesn't this make more sense, given the power of today's computers, and the limitations of all the analytic procedures? Which test of difference of variances is 'better' is likely to depend on the exact nature of the data....how non-normal is it? In what ways? With what sample sizes     

Group: comp.softsys.sas · Group Profile · Search for Computing The Variance Of Grouped Data in comp.softsys.sas
Author: BILL WEST
Date: Jan 24, 2008 06:17

On Sep 1, 2:11 am, Alan <jalantho...@verizon.net> wrote: Ray,      I am trying to determine the effect, if any, of two factors --- Message Size and Input Message Rate --- on the amount of time taken by software to process messages.  I used a 2-level factorial design for the experiment with 2 replications as a start, because this is our initial look at the data.     Timetags
Show full article (2.76Kb)
We are still unable to calulate the amount of space required with calulcation provided by MS. Can we get an update on this issue? "Paul Bruff" wrote: > Just to correct the post above: > > "So even with a maximum amount of 31GB used the calculation gives us just > over 40GB for the replica volume." > > - I mean recovery point volume not replica. > > > "Paul Bruff" wrote     

Group: comp.softsys.sas · Group Profile · Search for Computing The Variance Of Grouped Data in comp.softsys.sas
Author: Peter Flom
Date: Jan 3, 2008 19:38

Just to correct the post above: "So even with a maximum amount of 31GB used the calculation gives us just over 40GB for the replica volume." - I mean recovery point volume not replica. "Paul Bruff" wrote: Hi Madhan, 1. I have been looking at this with Rob, our monitoring system charts utilization of the disks over time. Since we have DPM installed on the system the
Show full article (5.06Kb)
Hi Madhan, 1. I have been looking at this with Rob, our monitoring system charts utilization of the disks over time. Since we have DPM installed on the system the amount of disk space used on the HDD varies between 22 - 31GB max used (week end file dumps get written, backed up to a second system and then removed). So even with a maximum amount of 31GB used the calculation gives us just     

Group: sci.stat.math · Group Profile · Search for Computing The Variance Of Grouped Data in sci.stat.math
Author: Robert Jones
Date: Sep 1, 2008 03:33

Ok, I actually simplified the problem somewhat. The actual model statement is model result = baseline dose; where baseline is a continuous variable. I didn't thought it would alter the solution to the the "dose problem" I described. I apologies if this added to confusion, but I'm thankful for the help I've received. Also, taking the baseline variable into account, it's now clear to me that
Show full article (3.15Kb) · Show article thread
I'm stumped. I simulated this question and was unable to reproduce the problem. Nor does it make sense from my understanding of how the software works, not if the real code is that simple. The only thing that comes to mind is differential weighting of the data with and without the where statement, but I cannot reproduce that with the code presented. Warren Schlechte -----Original Message     

Group: microsoft.public.dataprotectionmanager · Group Profile · Search for Computing The Variance Of Grouped Data in microsoft.public.dataprotectionmanager
Author: Rob
Date: Aug 21, 2008 05:05

Typically the point estimate of the LSMEANS and their difference (also ESTIMATE results) should be the same under the two fixed effects model scenarios you presented. It may help to see additional output or details of the data to determine what discrepancies you are actually facing. Robin High UNMC Joakim Englund <joakim.englund@GMAIL.COM> Sent by: "SAS(r) Discussion" <SAS-L@LISTSERV
Show full article (6.72Kb) · Show article thread
    

Group: microsoft.public.dataprotectionmanager · Group Profile · Search for Computing The Variance Of Grouped Data in microsoft.public.dataprotectionmanager
Author: Paul Bruff
Date: Aug 15, 2008 03:56

Show full article (6.22Kb) · Show article thread
    

Group: microsoft.public.dataprotectionmanager · Group Profile · Search for Computing The Variance Of Grouped Data in microsoft.public.dataprotectionmanager
Author: Paul Bruff
Date: Aug 15, 2008 02:24

Show full article (5.68Kb) · Show article thread
    

Group: comp.softsys.sas · Group Profile · Search for Computing The Variance Of Grouped Data in comp.softsys.sas
Author: Joakim Englund
Date: Aug 1, 2008 07:34

Show full article (7.80Kb)
    

Group: comp.softsys.sas · Group Profile · Search for Computing The Variance Of Grouped Data in comp.softsys.sas
Author: Warren Schlechte
Date: Aug 1, 2008 07:09

Show full article (6.63Kb)
    

Group: comp.softsys.sas · Group Profile · Search for Computing The Variance Of Grouped Data in comp.softsys.sas
Author: Robin R High
Date: Jul 30, 2008 07:11

Show full article (6.66Kb)
1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · next