www.cmm.bristol.ac.uk/forum

I have been working with Stat-JR:TREE successfully using the manual to learn it. However, sometimes on opening the program, and sometimes during the running of the program in the browser, I get "local host not responding" and things grind to a halt such that I usually have to start over. Any idea what might be going on or how I can fix this? Will Shadish

Would it be possible to provide more information about the computer that you are running it on, in particular the version of the browser that you are using? It would also be useful if you could provide screenshots of the browser and console windows at the point where this happens, as well as describe your actions prior to it freezing up if there is a patten as to when this occurs.

Once thing to check for is whether the problem coincides with your computer changing network state, i.e. leaving/joining a network or VPN.

Chris, I am using an HP 64bit PC. I am using IE 11.0.9600. Sequence is I start Stat-JR: TREE from the start menu, the dos screen opens, then IE opens. I move the cursor to either the dataset or template icon and left click, and that is when things stop. I have attached four snapshots: (a) dos screen 1, (2) dos screen 2 that continues dos screen 1 since I could not fit the entire dos screen in one snapshot, (3) screen freeze snapshot that shows what IE looks like when it freezes (note template is white as I clicked on it, but no menu came down; note also the local host message lower left of IE), and (4) a subsequent snapshot of IE after a "local host is not responding" error bar at the bottom of IE--whoops, I see a message above saying ""Cannot add another attachment, 3 is the maximum", so I'll send another email with the fourth snapshot. I am running this on my work computer at UC Merced, which is connected to the UCM network. I have not actively used VPN to get on the network, though I do not know if some sort of VPN is set up by our IT folks in the background. I can find out if needed. I had a milder version of this problem from the start, but the program seemed able to recover and I successfully used it to replicate the material in the beginners manual through p 60. But now, I can't do anything. It just freezes at startup. Thanks. Will Shadish

Chris, the fourth snapshot is attached. Will

Thanks for the additional information, we are looking into whether we are able to replicate this behaviour. In the mean time would it be possible for you to try this with a different browser (we mostly developed and tested the software with Firefox and Chrome) to see whether you get similar behaviour there?

So far Google Chrome is working fine. I'll continue and report what happens.

In the screen snapshot on p. 65 of the beginners manual, middle snapshot shows expression age^age. That returns an error. But age*age works.

Will

I'm now into the bang1 example that goes through p 69 of the beginner's manual. I'm getting different results. I've got five snapshots to send you, three attached. Back in a sec. Will

Here are the other two. The p 66 snapshot shows what seems to be identical choices as the two screen snapshots on p 66 of the manual. The Model.txt snapshot shows what seems to be identical model text on the top snapshot of p. 67 of the manual. However, the tau_u.xml snapshot shows a different model for the second equation. The other two snapshots continue to show different results. I've triple-checked my input and think I've done it right, but the snapshots in the manual have quite small text that perhaps I am misreading.

I started to do the OpenBugs example starting page 69. The snapshot shows eStat as the estimation engine when I think you intended it to be OpenBugs. In any case, if I run it using OpenBUGS, I get the error attached (openbugs.jpg) and the program stops executing after only 3s.

By the way, the popout menu scroll bar does not work properly. If I move up and down the menu list with arrow keys, all is fine. If I try to use the scroll bar, it will not scroll down below the last menu item that happens to be visible, and will simply open the menu item that happens to be where I place my cursor and right click on the scoll bar.

I hope I am not being a pain in the neck. If somehow these are my errors I would be mortified. But I am hoping that I am being helpful.

Will

Thank you for the very detailed and helpful feedback, I will attempt to address it below.

• If you zoom in enough on the picture on page 65 of the beginner's manual you will see that the expression is age*age, however due to the size of the picture this is hard to see. We will look into making this more obvious in a later edition. Incidentally for a full list of expressions supported by this screen see https://code.google.com/p/numexpr/wiki/ ... _operators.
• From your first screenshot relating to page 69 of the beginner's manual it appears that you have expression optimisation turned off. To enable this click the "settings" button on the top right of the window and tick the "Optimise generated code" checkbox the window this opens. Finally click "Set" at the bottom of the page. Do you remember changing any of these settings, as this should be on by default? If not we will investigate why this might have changed. I suspect that this will also explain the slight differences in results that you are seeing, as reorganising the expressions could results in slightly different answers due to rounding errors, etc.
• The OpenBUGS error is a bit puzzling. Two possibilities that I can think of are that you are using a different version of OpenBUGS (we tested with 3.2.2) or that it is having problems reading/writing to the temporary directory. Are there any messages in the console window corresponding to when this occurs?
• I will attempt to replicate the scrollbar issue that you were seeing. Does this problem occur with all the templates that you have tried?

1. OK on ^ vs *. It was hard to read.
2. I turned on expression optimization. I do not recall turning it off. I had gone into settings to indicate where the various programs were on my c: drive, and perhaps I somehow inadvertently changed it then. I will check if this solves the numerical results tomorrow.
3. I'm using OpenBUGS 3.2.1 Rev 781. Sometime this week I will upgrade it and see if that works.
4. I will keep track if the scrollbar issue occurs with other templates. I only noticed it that once.

I'm a bit swamped today, but will try to get back to verifying numerical results and answering your other questions tomorrow.

Thanks so much for your help.

Will