Simulation executes 2x as many times as expected

[antoniohramirez]

It looks like basic_simulation.json runs twice as many times as we expected.

We have "number_of_runs": 2 in the config file and have the mTree runner set to -m, but we have 4 outputs in the log file.

Why does this happen?

[skunath]

Can you send the relevant log output?

I'd check to see if you have any other references to the experiment log to make sure there's no additional outputs.

On Tue, Jul 23, 2019 at 7:21 PM antoniohramirez notifications@github.com wrote:

It looks like basic_simulation.json runs twice as many times as we expected.

We have "number_of_runs": 2 in the config file and have the mTree runner set to -m, but we have 4 outputs in the log file.

Why does this happen?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3F5NJGYQIHFP2AO52DQA5K45A5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HBAM4IQ, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3B37ACLSZTZNVE2UH3QA5K45ANCNFSM4IGIENYA .

[antoniohramirez]

mtree.log

[skunath]

Which branch is this code on?

On Tue, Jul 23, 2019 at 7:30 PM antoniohramirez notifications@github.com wrote:

mtree.log https://github.com/gmucsn/mTree/files/3423772/mtree.log

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3HILTTHBJ5YBXB544TQA5L6VA5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD2UF7RA#issuecomment-514351044, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3EECHBUSINJ2MP5GSLQA5L6VANCNFSM4IGIENYA .

[antoniohramirez]

We just pushed to master after making some structural changes to our agent strategy classes. It looks like we don't have the running 2x problem anymore, but now it only runs 1-2 more times than we expect. In simulation_launch.json, we expect 12 runs, but the mtree.log gives us 13-14 outputs

[skunath]

OK. So you have like a baker's dozen sims? Isn't that a better value?

I'll pull the code when I get home and see if there is any problem with the sim dispatcher.

On Thu, Jul 25, 2019 at 11:47 AM antoniohramirez notifications@github.com wrote:

We just pushed to master after making some structural changes to our agent strategy classes. It looks like we don't have the running 2x problem anymore, but now it only runs 1-2 more times than we expect. In simulation_launch.json, we expect 12 runs, but the mtree.log gives us 13-14 outputs

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3HIP2N7T3MU54KEALLQBHDJJA5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD2Z4NHA#issuecomment-515098268, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3EWID3AAWYX7V2FNSDQBHDJJANCNFSM4IGIENYA .

[antoniohramirez]

Did you find anything in that sim dispatcher?

[skunath]

So so far I've not been able to replicate your numbers. In the current config I only get three outputs which would seem to be as expected.

I've almost got the simulation_id and run_number problem fixed. Hopefully I can get a patch before noon.

On Fri, Jul 26, 2019 at 10:54 AM antoniohramirez notifications@github.com wrote:

Did you find anything in that sim dispatcher?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3BXRCXLIV4W73TRQNLQBMFYRA5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD24256Y#issuecomment-515485435, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3HI4L6PPYWSJUUBCLLQBMFYRANCNFSM4IGIENYA .

[apsurek]

We have tested that this issue (running multiple times) persists across both Mac and Windows. Maybe try running it with a higher number of runs and running multiple configs. We have created a config file called "erroring_setup.json" which should help you replicate the error. All of this code is on the master branch, so if you pull from there and run the config, you should see this error.

[skunath]

I'm still getting the right number -- 400.

Let me get the simulation_id and run_number id and maybe that could help debug yours.

On Mon, Jul 29, 2019 at 11:24 AM apsurek notifications@github.com wrote:

We have tested that this issue (running multiple times) persists across both Mac and Windows. Maybe try running it with a higher number of runs and running multiple configs. We have created a config file called "erroring_setup.json" which should help you replicate the error. All of this code is on the master branch, so if you pull from there and run the config, you should see this error.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3GRVQYA7VXNPZWIHUTQB4DSZA5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3BCAPA#issuecomment-516038716, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3GNPX4NREQBTW5ZCIDQB4DSZANCNFSM4IGIENYA .

[skunath]

Also, if you pull the latest updates to mTree you should get some new properties on the institution.

These two should probably be included in your log statements now. Hopefully that can help us track down the extra runs issue.

self.simulation_id self.run_number

On Mon, Jul 29, 2019 at 11:56 AM Stephen Kunath skunath@gmail.com wrote:

I'm still getting the right number -- 400.

Let me get the simulation_id and run_number id and maybe that could help debug yours.

On Mon, Jul 29, 2019 at 11:24 AM apsurek notifications@github.com wrote:

We have tested that this issue (running multiple times) persists across both Mac and Windows. Maybe try running it with a higher number of runs and running multiple configs. We have created a config file called "erroring_setup.json" which should help you replicate the error. All of this code is on the master branch, so if you pull from there and run the config, you should see this error.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3GRVQYA7VXNPZWIHUTQB4DSZA5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3BCAPA#issuecomment-516038716, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3GNPX4NREQBTW5ZCIDQB4DSZANCNFSM4IGIENYA .

[apsurek]

I just pulled mTree 3 seconds ago and there was no updated to pull.

[skunath]

Sorry. Forgot to do the pull request. Try again now.

On Mon, Jul 29, 2019 at 12:14 PM apsurek notifications@github.com wrote:

I just pulled mTree 3 seconds ago and there was no updated to pull.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/gmucsn/mTree/issues/33?email_source=notifications&email_token=AABDF3AMLZTHVE7UF2VW4STQB4JPFA5CNFSM4IGIENYKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3BGZ4A#issuecomment-516058352, or mute the thread https://github.com/notifications/unsubscribe-auth/AABDF3C7KBKMTIQ644UGM5DQB4JPFANCNFSM4IGIENYA .

[antoniohramirez]

Try erroring_setup_2.json. I just pushed it to master, and it ran 280 times when I specified 240

[skunath]

Will do.

Were you able to add the run number and simulation id to the log?

On Jul 29, 2019, at 1:38 PM, antoniohramirez notifications@github.com wrote:

Try erroring_setup_2.json. I just pushed it to master, and it ran 280 times when I specified 240

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[antoniohramirez]

Yes, the run_number and simualtion_id work -- it helped us debug the issue