Experiment Mismatch

Question

question

Dash Ballarta asked Dec 16 2016 at 9:48 PM Matthew Gillespie commented Dec 21 2016 at 8:46 PM

Experiment Mismatch

When I run a 1 day iteration, I'll get, for example, an output of 4/9/9/5 for PCIs 1-4, which is what I'm expecting.

However, when I run the experimenter, even for one iteration, I will recieve 0/2/2/4 or something wildly different from a one day in-model iteration. This will also be represented if I do a simulation of any magnitude. Wondering if you could help troubleshoot.

FSM is attached.

whv1.fsm

Software Version:

FlexSim HC 5.1.0

experimenter

whv1.fsm (6.1 MiB)

· 4

Dash Ballarta commented · Dec 20 2016 at 8:49 PM

@Matthew Gillespie

Check out this replication. I only ran one replication and these MA's are immediate off schedule and now cause immense queue since they can't do any duties. I have no settings (i think) that would result in them being this way.

Off of experimenter. No issues if run in normal mode. Is it possibly a time/date thing?

0 ·

untitled.png (445.3 KiB)

Matthew Gillespie ♦♦ Dash Ballarta commented · Dec 20 2016 at 9:07 PM

So you have an experiment variable that's setting the number of active resources on the MA group to 0. You must have deleted the name in the table though since it's not showing up in the table, but you can see it in the tree.

To fix it just click in the cell for variable 1 and reselect the empty option at the top of the list.

That's why I couldn't reproduce this at first because I clicked in that cell to see if there was an experiment variable being set.

0 ·

expvariable.png (7.9 KiB)

Dash Ballarta Matthew Gillespie ♦♦ commented · Dec 21 2016 at 8:43 PM

Hi @Matthew Gillespie, thanks for the various looks so far. I've made the changes and one of those clinics is running smoothly.

I've gone back and tried to clean up some logic for one of the other clinic spaces. In my PCI #5, I get this error:

time: 217.662807 exception: Exception Caught in Operator::onEndTask(treenode task) object: /WHRN_2 time: 217.662807 exception: Exception Caught in TaskExecuter::finishTask(treenode task) object: /WHRN_2 time: 217.662807 exception: Exception Caught in TaskExecuter::beginTask(treenode task) object: /WHRN_2 time: 217.662807 exception: Exception Caught in TaskExecuter::finishTask(treenode task) object: /WHRN_2 time: 217.662807 exception: Exception Caught in TaskExecuter::onTimerEvent(treenode involved, int code, char *datastr) object: /WHRN_2 time: 217.662807 exception: Exception Caught in ObjectFunction237__project_library_FlexSimObject_behaviour_eventfunctions_OnTimerEvent object: /WHRN_2 class: /WHRN_2

Doesnt seem to be affecting my simulation, but I don't know where the logic is corrupting along the way.

0 ·

Show more comments

Answer 1 · 2016-12-17T00:02:29Z

Matthew Gillespie answered Dec 17 2016 at 12:02 AM Matthew Gillespie commented Dec 20 2016 at 7:55 PM

In the experimenter if you go to the Explore tab you'll see a couple drop downs where you can pick a specific replication. This sets up your model so that you can run the model with all the same settings as that replication and see what happened in that replication. This is the easiest way to debug an individual replication like this. Note though, that there's no way to revert the model to its original state after doing this, so you might not want to save the model after doing this.

Doing the above I was able to see that a lot of your patients get stuck in WaitingArea1 and never leave. Most of them are in the middle of activity 50. It looks like these patients do activity 30 and then do activity 40 and 50 at the same time. I'm not sure why they're not moving, but it looks like the MA staff that are supposed to do the maintenance in the ExamTable beds are always off schedule.

· 9

Dash Ballarta commented · Dec 19 2016 at 3:18 PM

Matt,

Thanks for the look. I did what you recommended and see that the MAs are definitely off shift. I'm not sure why that's applying only to them and not to any other groups... I removed the staff schedules altogether and they are still no available.

I'll try to reset some of the parameters and see if that helps.

Dash

0 ·

Dash Ballarta Dash Ballarta commented · Dec 19 2016 at 3:30 PM

I applied some changes to the model and found that:

1) For some reason the MA staff group was defaulting to 0 active members, I was able to reset this and replicate a "Day 1" iteration.

2)My outputs after 100 replications still show low patient outputs, even though as I explore day by day, outputs seem reasonable across the board (e.g., i am expecting 5/7/7/5 or so, as I explore individually they are all +/- 2, but the overall experiment dashboard still shows 2/2/4/1)

0 ·

Matthew Gillespie ♦♦ Dash Ballarta commented · Dec 19 2016 at 5:00 PM

I would switch your Patient Output chart to be a Dot plot so you can see the values of each replication side by side. That might help you spot outliers more quickly.

0 ·

Show more comments

Dash Ballarta Dash Ballarta commented · Dec 19 2016 at 4:06 PM

@Matthew Gillespie

Just another question as well. I'd like to run some collection off patients with the "spanish == 1" label. is there an easy way to do this either through an export or in the dashboard itself?

0 ·

Matthew Gillespie ♦♦ Dash Ballarta commented · Dec 19 2016 at 5:03 PM

If you add the Spanish label to the PCI table and reset your model you should get Spanish in the drop down of the Custom Sample Sets popup.

0 ·

question