topic Re: Semi/CoVariance Cloud POPUPS in ArcGIS GeoStatistical Analyst Questions

Semi/CoVariance Cloud POPUPS

TedCronin — Thu, 25 Aug 2011 23:31:29 GMT

So, if I open up my subset, and I have more than 500 records (I have a 1000), i get the popup. If I change the attribute, I get another popup. If I change the attribute again, I get another popup, etc..... Why not give me one popup, and then after the first one, not give me additional popups, seems a bit redundant. I get it. i have more than 500 points, I don't need to be reminded over and over and over and over and over and over.

Re: Semi/CoVariance Cloud POPUPS

TedCronin — Thu, 25 Aug 2011 23:38:12 GMT

Perhaps a similiar behavior to the Crosscovariance tool, where I set the top data source, see the annoying popus twice, but when I hit the second tier, I don't see the popups.

It seems like I may have experienced a crash with this tool as well, not a physical crash, but a not responding crash.

Duped twice.

So, it completes, but I do get an unspecified error in the crosscovariance tab.

So, that's cool, after I get the unspecified error, I can still add it to layout, so that works... it is still unspecified, but I can see it on my map. if I shake the dialog, I can see the error jiggle 🙂

Re: Semi/CoVariance Cloud POPUPS

SteveLynch — Fri, 26 Aug 2011 15:05:22 GMT

Ted

These 2 ESDA tools creates "dots" on the graph, where each dot represents what happens between 2 points. The number of points in the graph can be calculated by computing n(n+1)/2, where n=number of points.

We have done testing and determined that n=500 is sufficient, i.e. 125,250 dots on the graph. Using more than 500 points really does not tell you much more, so we decided to warn you each time. n=1000 will create half a million dots, will you learn any more than if you used n=500?

A rule of thumb when doing ESDA is to create a subset (use the CreateSubsets GP tool) if you're using a large dataset. 1000 is large when talking to geostatisticians.

I cannot repro the crash.

Hope this helps and thanks for testing

Steve

Re: Semi/CoVariance Cloud POPUPS

TedCronin — Tue, 30 Aug 2011 20:56:40 GMT

Ted

These 2 ESDA tools creates "dots" on the graph, where each dot represents what happens between 2 points. The number of points in the graph can be calculated by computing n(n+1)/2, where n=number of points.

We have done testing and determined that n=500 is sufficient, i.e. 125,250 dots on the graph. Using more than 500 points really does not tell you much more, so we decided to warn you each time. n=1000 will create half a million dots, will you learn any more than if you used n=500?

A rule of thumb when doing ESDA is to create a subset (use the CreateSubsets GP tool) if you're using a large dataset. 1000 is large when talking to geostatisticians.

I cannot repro the crash.

Hope this helps and thanks for testing

Steve

Ok, I will retest on some other data, and see if I can repro. Are you guys using Win 7?

Re: Semi/CoVariance Cloud POPUPS

EricKrause — Thu, 01 Sep 2011 15:02:25 GMT

We have Win 7 and XP machines. I tested both and didn't get the crash.

Re: Semi/CoVariance Cloud POPUPS

TedCronin — Thu, 01 Sep 2011 15:07:46 GMT

We have Win 7 and XP machines. I tested both and didn't get the crash.

Thank you Eric. I am currently trying to automate a process for Sales Extraction, and once that has been completed, I will take a look again but with different data, if there are any crashes with this other data, I will attach or email directly. Not getting a crash on your end is a good thing, IMO, but will retest soon, anyway. this other layer I will be extracting out points and polys so I can go through your tutorial.

Re: Semi/CoVariance Cloud POPUPS

EricKrause — Fri, 02 Sep 2011 19:56:23 GMT

Even if it only happens with one dataset, we'd still like to know. If it's happening with one, it will happen with others.

Re: Semi/CoVariance Cloud POPUPS

TedCronin — Mon, 12 Sep 2011 21:13:33 GMT

ok, I will try and get a dataset out this week, hopefully by tomorrow.

Re: Semi/CoVariance Cloud POPUPS

TedCronin — Mon, 26 Sep 2011 18:13:48 GMT

I am back, now trying to rework my steps to see if I can repro anything. Had a blue screen from spatial stats on production machine, so I was a bit tied up.