Error/Bug in VA with visualization and crosstab

[tsjwolf]

Good day.
 
I have bug in VA with visualization and crosstab.
 
This is gameplay.
 
I use file "INSIGHT_TOY_COMPANY_2012", which is sample file with VA.
 
I have added data from Customer to "Facility Hierarchy" that I can see which "Customers" are working with Single Facility.

I copy Category "Customer"  to "Customer Geo".

I chose "Category" for "Customer Geo" from 

"Geography" -> "Custom",

"Latitude" -> "xyCustomer Lat",

"Longitude" -> "xyCustomer Lon", "

Coordinate Space "->" World Geodetic System (WGS84) "
 
Then in "Facility Hierarchy" I added CustomerGeo.
 
In the visualization "Facility Analysis" in "Facility Hierarchy" I have chosen "North America" -> "United States" -> "Montana" -> Billings and I see that there's in Billings various "FacilityEfficiency" Different "bubble size".

Foto in file VA1.png. 

With Button „Export Data“ wir get Excel-CSV-Data:

Customer Geo	Facility Age	Facility Efficiency
USBILLIN000190392	11	84.9999999999997300000000000000000%
USBILLIN000190391	11	84.9999999999997600000000000000000%
USBILLIN000190162	11	84.9999999999998400000000000000000%
USBILLIN000186984	11	84.9999999999998600000000000000000%
USBILLIN000186983	11	84.9999999999999100000000000000000%
USBILLIN000186988	11	84.9999999999999300000000000000000%
USBILLIN000186987	11	84.9999999999999700000000000000000%
USBILLIN000190161	11	84.9999999999999700000000000000000%
USBILLIN000191072	11	84.9999999999999700000000000000000%
USBILLIN000186990	11	85.0000000000000000000000000000000%
USBILLIN000190159	11	85.0000000000000000000000000000000%
USBILLIN000190163	11	85.0000000000000000000000000000000%
USBILLIN000191074	11	85.0000000000000100000000000000000%
USBILLIN000191070	11	85.0000000000000400000000000000000%
USBILLIN000191071	11	85.0000000000000700000000000000000%
USBILLIN000191073	11	85.0000000000000700000000000000000%

With „Cross Table“ I get Data:

Foto in file VA2.png. 

But I have tested original file of "INSIGHT_TOY_COMPANY_2012" with SAS.

PROC SQL;

   CREATE TABLE WORK.QUERY_FOR_INSIGHT_TOY_COMPA_0002 AS

   SELECT t1.Customer,

            (AVG(t1.FacilityAge)) FORMAT=NLNUMI3. AS AVG_of_FacilityAge,

            (AVG(t1.FacilityEfficiency)) FORMAT=PERCENT32.31 AS AVG_of_FacilityEfficiency,

            (MIN(t1.FacilityEfficiency)) FORMAT=PERCENT32.31 AS MIN_of_FacilityEfficiency,

            (MAX(t1.FacilityEfficiency)) FORMAT=PERCENT32.31 AS MAX_of_FacilityEfficiency,

            (COUNT(t1.Customer)) AS COUNT_of_Customer

      FROM DPPUBLIC.INSIGHT_TOY_COMPANY_2012 t1

      WHERE t1.Customer IN

           (

'USBILLIN000190391',

'USBILLIN000190392',

'USBILLIN000186984',

'USBILLIN000190162',

'USBILLIN000186983',

'USBILLIN000186987',

'USBILLIN000186988',

'USBILLIN000190161',

'USBILLIN000191072',

'USBILLIN000186990',

'USBILLIN000190159',

'USBILLIN000190163',

'USBILLIN000191070',

'USBILLIN000191071',

'USBILLIN000191073',

'USBILLIN000191074'           )

      GROUP BY t1.Customer

                  ORDER BY t1.Customer

;

QUIT;

All these facilities have the same "Facility Efficiency", which is 85%

Customer	AVG_of_FacilityAge	AVG_of_FacilityEfficiency	MIN_of_FacilityEfficiency	MAX_of_FacilityEfficiency	COUNT_of_Customer
USBILLIN000186983	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	534
USBILLIN000186984	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	523
USBILLIN000186987	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	409
USBILLIN000186988	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	488
USBILLIN000186990	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	400
USBILLIN000190159	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	23
USBILLIN000190161	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	16
USBILLIN000190162	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	121
USBILLIN000190163	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	22
USBILLIN000190391	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	442
USBILLIN000190392	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	519
USBILLIN000191070	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	81
USBILLIN000191071	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	58
USBILLIN000191072	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	12
USBILLIN000191073	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	58
USBILLIN000191074	11	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	85,0000000000000000000000000000000%	25

VA3.png and VA4.png is the photos, how to make "CustomerGeo".

I think maybe there's error or problem or bug in the calculation in VA.

I do not know how to find solution.

[AnnaBrown]

Hi  tsjwolf,

Thanks for posting your situation to the SAS Visual Analytics Community! I checked on this and SAS Technical Support is actively working on a solution and will continue to be in touch on their progress.

Many thanks,

Anna

[boemskats]

I just came here from your LinkedIn post. Just to be clear - is the error you describe in your LI post as a 'gross arithmetical error' the loss of precision after the 12th decimal, in cases when there is not a format applied to the result? If so, this just looks like a floating point issue to me.

I'm just trying to understand whether I'm missing something obvious, as I'm not a statistician.

Nik