BookmarkSubscribeRSS Feed
Noel88
Obsidian | Level 7

I'm not having much luck trying to work out how to take a data series and create a XmR control chart in VA. I can easily do this in excel. Here's a typical data set.

If you treat the columns as A - E in an excel spreadsheet.

 

Number clients is the straight data which updates each month.
Average is the average of the column of clients over the period of the data set.  =AVERAGE(B:B) in excel.
Range is the difference between one month and the previous =ABS(B6-B7) in excel

Range average is well the range average =AVERAGE(F:F) in excel

LCL is the lower limit calculation =C6-(2.66*G6) in excel

UCL is the upper limit calculation  =C6+(2.66*G6) in excel

While I may have to change the upper and lower limits if on occasion trends continue for most of the subsets that fall off this data they do stay within the upper and lower limits most of the time. As the client number data drops down from a national, to a regional and then a local  level obviously it is preferable that the calculations are dynamic. I could get away with fixed lines for the average and lower limits if that was the only way to do this - I'd just have to do multiple graphs. The move from excel and manual updating to VA and auto updating the client numbers outweighs the intial effort to do this.

I only have access to VA and not to building data tables that could do these calculations at the back end.

 

MonthNumber clientsAverageLCLUCLRangeRange average
Jul-1622,43322,264   21,806   22,722 172
Aug-1622,61222,264   21,806   22,722179172
Sep-1622,66022,264   21,806   22,72248172
Oct-1622,38022,264   21,806   22,722280172
Nov-1622,54522,264   21,806   22,722165172
Dec-1622,90322,264   21,806   22,722358172
Jan-1722,84322,264   21,806   22,72260172
Feb-1722,59522,264   21,806   22,722248172
Mar-1722,07822,264   21,806   22,722517172
Apr-1721,94222,264   21,806   22,722136172
May-1721,85522,264   21,806   22,72287172
Jun-1721,72022,264   21,806   22,722135172
Jul-1721,57822,264   21,806   22,722142172
Aug-1721,55122,264   21,806   22,72227172
Sep-1721,50122,264   21,806   22,72250172
Oct-1721,45122,264   21,806   22,72250172
Nov-1721,64622,264   21,806   22,722195172
Dec-1722,35222,264   21,806   22,722706172
Jan-1822,46022,264   21,806   22,722108172
Feb-1822,35522,264   21,806   22,722105172
Mar-1822,10222,264   21,806   22,722253172
Apr-1822,13322,264   21,806   22,72231172
May-1822,02822,264   21,806   22,722105172
Jun-1822,10922,264   21,806   22,72281172
Jul-1822,02422,264   21,806   22,72285172
Aug-1822,18822,264   21,806   22,722164172
Sep-1822,23022,264   21,806   22,72242172
Oct-1822,03422,264   21,806   22,722196172
Nov-1822,05922,264   21,806   22,72225172
Dec-1822,77622,264   21,806   22,722717172
Jan-1922,89222,264   21,806   22,722116172
Feb-1922,70322,264   21,806   22,722189172
Mar-1922,47222,264   21,806   22,722231172
Apr-1922,26522,264   21,806   22,722207172
May-1922,34522,264   21,806   22,72280172
Jun-1922,45422,264   21,806   22,722109172
Jul-1922,33422,264   21,806   22,722120172
Aug-1922,57522,264   21,806   22,722241172
Sep-1922,68222,264   21,806   22,722107172
Oct-1922,69622,264   21,806   22,72214172

 

Any help or thoughts would be appreciated.

It would be quite cool down the track to have a specific graph option for these type of control charts with the option to modify the 2.66 calculation in the limit workings  =C6-(2.66*G6). The other calculations are quite standard.

8 REPLIES 8
Frank_Boekamp
Quartz | Level 8

In SAS VA 7.4 I don't know but in SAS Viya proc scp is now available in CAS, creating the tables for XMR charts. See

Process Monitoring in SAS Viya with the SPC Procedure https://www.youtube.com/watch?v=HNmJicNznIg

 

A visual component you have to build yourself though.It would be great to see a visual object in SAS VA someday to create XMR charts OOTB.

Noel88
Obsidian | Level 7

Unfortunately we do not have Viya at work. Nice to know it has been thought about though.

Trying to move people's thinking away from targets and the often negative behaviour and practice  that drives. Being able to show an alternate way of thinking via reporting is a crucial part of that.

Noel88
Obsidian | Level 7

Been trying to do that for 20 years unfortunately. The recent move to VA for reporting in our organisation has democratised the use of data. The ability to build reports locally that can be utilised and viewed nationally is driving some good change in how reporting is viewed and used and influences. 

Having views that don't come from the centre is making people think.

Noel88
Obsidian | Level 7

Just wondering if anyone can come up for a solution to this. It seems weird to me that something I can easily do in excel I can't figure out here. I was thinking it was me not quite getting how to fully utilise VA but maybe not.

SASKiwi
PROC Star

@Noel88  - You could raise this as a SASware Ballot suggestion: https://communities.sas.com/t5/SASware-Ballot-Ideas/idb-p/sas_ideas

 

Any improvements would only be added to the latest version of VA though.

Noel88
Obsidian | Level 7

Cool. Thanks for letting me know.

Noel88
Obsidian | Level 7

Managed to work out a work around where I can calculate the mean and limits in excel and then add these to the graph as reference lines. Means while the data updates each month automatically I need to manually make any adjustments to the mean, UCL and LCL.

Works as an interim solution till I find something better and gives the report visibility needed throughout the monthVA Control chart.JPGVA Control chart 2.JPG

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

Tips for filtering data sources in SAS Visual Analytics

See how to use one filter for multiple data sources by mapping your data from SAS’ Alexandria McCall.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 8 replies
  • 3193 views
  • 1 like
  • 3 in conversation