Container?

[gailrosen]

Hi,

Do you have a containerized version of COMETS? I am trying to go through the hello.world example and finding that things are almost deprecated. (or at least the way that cobra imports a model has changed)

It might be best to just containerize this to help novices use it. Do you have a Docker? singularity file?

thanks,
Gail

[dukovski]

Yes, cobra replaced cobra.test with cobra.io. That's it. Everything else should work OK.
And yes, we are working on a container. I hope to do it by the end of the February.

[gailrosen]

I think a container that "just works" (at a point in time) would be GREAT.. Thank you!

[dukovski]

Agree! Would you help me do it?

[gailrosen]

I'm not really qualified. I still can't get COMETS to work. I am still getting "Traceback (most recent call last): File "/Users/gailrosen/Documents/engr132_winter25/lecture/lecture_code/testcomets.py", line 23, in <module> sim.run() ~~~~~~~^^ File "/Library/Frameworks/Python.framework/Versions/3.13/lib/python3.13/site-packages/cometspy/comets.py", line 401, in run self.__analyze_run_output() ~~~~~~~~~~~~~~~~~~~~~~~~~^^ File "/Library/Frameworks/Python.framework/Versions/3.13/lib/python3.13/site-packages/cometspy/comets.py", line 566, in __analyze_run_output raise RuntimeError(f"COMETS simulation did not complete:\n {message}") RuntimeError: COMETS simulation did not complete: JAVA could not find gurobi. try the following:

>> import os >> os.environ['GUROBI_COMETS_HOME']

if there is nothing there try setting that variable to the location of gurobi.jar, for example:

>> os.environ['GUROBI_COMETS_HOME'] = '/opt/gurobi900/linux64'"

AND I have pointed GUROBI_COMETS_HOME directly to the gurobi jar with: os.environ['GUROBI_COMETS_HOME'] = '/Library/gurobi1200/macos_universal2/lib' So, I have no idea why I can't get it to work. I look forward to the container where I don't have to struggle with all this setup! thanks! Gail

…

________________________________________ From: dukovski ***@***.***> Sent: Friday, January 17, 2025 9:43 AM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Agree! Would you help me do it? — Reply to this email directly, view it on GitHub<#54 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABHZROI7AG3PVN6TPAWDSW32LEJHLAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJYGUYTQNZYG4>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[dukovski]

Are on a Mac by any chance. It seems that Macs have problems with Gurobi now. I have OR-Tools implemented too. I will fix this early next week.

[gailrosen]

Yes -- I have a mac. Should I use windows? or can i get away with a linux version?

…

________________________________________ From: dukovski ***@***.***> Sent: Saturday, January 18, 2025 7:37 AM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Are on a Mac by any chance. It seems that Macs have problems with Gurobi now. I have OR-Tools implemented too. I will fix this early next week. — Reply to this email directly, view it on GitHub<#54 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABHZROO6YUMVJRC7KRXQEBD2LJDJVAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJZGY4TSOJUGU>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[dukovski]

Yes, only Mac gives me trouble lately. Try any of the other systems.

[gailrosen]

Hi, I have found that I get the same issue on Windows... I point the directory to the directory with gurobi.jar, but it doesn't see it. See below: sim = c.comets(ly, p) sim.run() Running COMETS simulation ... Debug Here ... Error: COMETS simulation did not complete examine comets.run_output for the full java trace if we detect a common reason, it will be stated in the RuntimeError at the bottom

--------------------------------------------------------------------------- RuntimeError Traceback (most recent call last) Cell In[33], line 2 1 sim = c.comets(ly, p) ----> 2 sim.run() File ~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.13_qbz5n2kfra8p0\LocalCache\local-packages\Python313\site-packages\cometspy\comets.py:401, in comets.run(self, delete_files) 398 self.run_errors = "STDERR empty." 400 # Raise RuntimeError if simulation had nonzero exit --> 401 self.__analyze_run_output() 403 # '''----------- READ OUTPUT ---------------------------------------''' 404 # Read total biomass output 405 if self.parameters.all_params['writeTotalBiomassLog']: File ~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.13_qbz5n2kfra8p0\LocalCache\local-packages\Python313\site-packages\cometspy\comets.py:566, in comets.__analyze_run_output(self) 564 message += "if in Unix. In Windows, it suggests that something changed\n" 565 message += "with the dependencies installed alongside COMETS" --> 566 raise RuntimeError(f"COMETS simulation did not complete:\n {message}") 568 message = "undetected reason. examine comets.run_output for JAVA trace" 569 raise RuntimeError(f"COMETS simulation did not complete:\n {message}") RuntimeError: COMETS simulation did not complete: JAVA could not find gurobi. try the following:

>> import os >> os.environ['GUROBI_COMETS_HOME']

if there is nothing there try setting that variable to the location of gurobi.jar, for example:

>> os.environ['GUROBI_COMETS_HOME'] = '/opt/gurobi900/linux64'

os.environ['GUROBI_COMETS_HOME'] 'C:\\gurobi1200\\win64\\lib' dir C:\\gurobi1200\\win64\\lib Directory: C:\gurobi1200\win64\lib Mode LastWriteTime Length Name ---- ------------- ------ ---- -a---- 11/3/2024 5:02 PM 282508 gurobi-javadoc.jar -a---- 11/3/2024 5:02 PM 150250 gurobi.jar -a---- 11/3/2024 5:02 PM 805 gurobi.py -a---- 11/3/2024 5:02 PM 55290 gurobi120.lib -a---- 11/3/2024 5:31 PM 171480 gurobi120.netstandard20.dll -a---- 11/3/2024 5:02 PM 504265 gurobi120.netstandard20.xml -a---- 11/3/2024 5:02 PM 2348990 gurobi_c++md2017.lib -a---- 11/3/2024 5:02 PM 2945264 gurobi_c++mdd2017.lib -a---- 11/3/2024 5:02 PM 2879514 gurobi_c++mt2017.lib -a---- 11/3/2024 5:02 PM 3517408 gurobi_c++mtd2017.lib -a---- 11/3/2024 5:02 PM 1215 rootcert.pem

…

________________________________________ From: dukovski ***@***.***> Sent: Saturday, January 18, 2025 1:30 PM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Yes, only Mac gives me trouble lately. Try any of the other systems. — Reply to this email directly, view it on GitHub<https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fsegrelab%2Fcomets%2Fissues%2F54%23issuecomment-2599821296&data=05%7C02%7Cglr26%40drexel.edu%7C5670edf368ec4fb98f2f08dd37ee313c%7C3664e6fa47bd45a696708c4f080f8ca6%7C0%7C0%7C638728218370600067%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=QH4Ih%2FbweybZRPpl74xnSmI4UyAPj6gPU%2BI9%2B8CH6Pw%3D&reserved=0>, or unsubscribe<https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FABHZROPQDVIAPG2Z2IRKTY32LKMUPAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJZHAZDCMRZGY&data=05%7C02%7Cglr26%40drexel.edu%7C5670edf368ec4fb98f2f08dd37ee313c%7C3664e6fa47bd45a696708c4f080f8ca6%7C0%7C0%7C638728218370618867%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=SmPwEhK8W%2BXuzqe%2BXZjwPlMBdWxode9W4xAn0PNnA%2F8%3D&reserved=0>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[gailrosen]

I'm back on the mac since I discovered that I just had to downgrade to Gurobi 10 Again.. it would save a lot of time and effort if we just had a container so i don't have to downgrade things and install packages.. it has consumed so much of my time. I still get an error: my_cmap = matplotlib.cm.get_cmap("copper") Traceback (most recent call last): File "testcomets.py", line 108, in <module> im = sim.get_biomass_image('e_coli_core', p.all_params["maxCycles"]) File "/Library/Frameworks/Python.framework/Versions/3.13/lib/python3.13/site-packages/cometspy/comets.py", line 662, in get_biomass_image if model_id not in list(np.unique(self.biomass['species'])): ^^^^^^^^^^^^ AttributeError: 'comets' object has no attribute 'biomass' I don't know why

________________________________________ From: Rosen,Gail ***@***.***> Sent: Wednesday, January 22, 2025 12:55 PM To: segrelab/comets; segrelab/comets Cc: Author Subject: Re: [segrelab/comets] Container? (Issue #54) Hi, I have found that I get the same issue on Windows... I point the directory to the directory with gurobi.jar, but it doesn't see it. See below: sim = c.comets(ly, p) sim.run() Running COMETS simulation ... Debug Here ... Error: COMETS simulation did not complete examine comets.run_output for the full java trace if we detect a common reason, it will be stated in the RuntimeError at the bottom

--------------------------------------------------------------------------- RuntimeError Traceback (most recent call last) Cell In[33], line 2 1 sim = c.comets(ly, p) ----> 2 sim.run() File ~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.13_qbz5n2kfra8p0\LocalCache\local-packages\Python313\site-packages\cometspy\comets.py:401, in comets.run(self, delete_files) 398 self.run_errors = "STDERR empty." 400 # Raise RuntimeError if simulation had nonzero exit --> 401 self.__analyze_run_output() 403 # '''----------- READ OUTPUT ---------------------------------------''' 404 # Read total biomass output 405 if self.parameters.all_params['writeTotalBiomassLog']: File ~\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.13_qbz5n2kfra8p0\LocalCache\local-packages\Python313\site-packages\cometspy\comets.py:566, in comets.__analyze_run_output(self) 564 message += "if in Unix. In Windows, it suggests that something changed\n" 565 message += "with the dependencies installed alongside COMETS" --> 566 raise RuntimeError(f"COMETS simulation did not complete:\n {message}") 568 message = "undetected reason. examine comets.run_output for JAVA trace" 569 raise RuntimeError(f"COMETS simulation did not complete:\n {message}") RuntimeError: COMETS simulation did not complete: JAVA could not find gurobi. try the following:

>> import os >> os.environ['GUROBI_COMETS_HOME']

if there is nothing there try setting that variable to the location of gurobi.jar, for example:

>> os.environ['GUROBI_COMETS_HOME'] = '/opt/gurobi900/linux64'

os.environ['GUROBI_COMETS_HOME'] 'C:\\gurobi1200\\win64\\lib' dir C:\\gurobi1200\\win64\\lib Directory: C:\gurobi1200\win64\lib Mode LastWriteTime Length Name ---- ------------- ------ ---- -a---- 11/3/2024 5:02 PM 282508 gurobi-javadoc.jar -a---- 11/3/2024 5:02 PM 150250 gurobi.jar -a---- 11/3/2024 5:02 PM 805 gurobi.py -a---- 11/3/2024 5:02 PM 55290 gurobi120.lib -a---- 11/3/2024 5:31 PM 171480 gurobi120.netstandard20.dll -a---- 11/3/2024 5:02 PM 504265 gurobi120.netstandard20.xml -a---- 11/3/2024 5:02 PM 2348990 gurobi_c++md2017.lib -a---- 11/3/2024 5:02 PM 2945264 gurobi_c++mdd2017.lib -a---- 11/3/2024 5:02 PM 2879514 gurobi_c++mt2017.lib -a---- 11/3/2024 5:02 PM 3517408 gurobi_c++mtd2017.lib -a---- 11/3/2024 5:02 PM 1215 rootcert.pem

…

________________________________________ From: dukovski ***@***.***> Sent: Saturday, January 18, 2025 1:30 PM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Yes, only Mac gives me trouble lately. Try any of the other systems. — Reply to this email directly, view it on GitHub<#54 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABHZROPQDVIAPG2Z2IRKTY32LKMUPAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJZHAZDCMRZGY>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[dukovski]

Yes, I agree. I will make a container.
Thanks for letting me know. Gurobi keep making changes
that is hard to keep up with. I have a version with OR-tools
that I will release soon, which will help deal with these kind of problems.

Thanks!

[gailrosen]

thanks.. i am trying to get this to work in the next week. Do you know why I am getting the "AttributeError: 'comets' object has no attribute 'biomass'" ? Anyway I can do a quick fix for this? is cometspy not loading properly -- any ideas?

…

________________________________________ From: dukovski ***@***.***> Sent: Thursday, January 23, 2025 8:03 AM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Yes, I agree. I will make a container. Thanks for letting me know. Gurobi keep making changes that is hard to keep up with. I have a version with OR-tools that I will release soon, which will help deal with these kind of problems. Thanks! — Reply to this email directly, view it on GitHub<#54 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABHZROMVM5VJPSLD2AH4T6D2MDSCLAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMBZG42TMNBTGI>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[dukovski]

Yes, you need to set the parameter to save the biomass. The total_biomass is saved by default, but that is the integrated biomass. If you want to save the spatial distribution of the biomass you need to specify:
p = cometspy.params()
p.set_param("writeBiomassLog", True)

also set the rate of writing the log.

I hope this helps.

[gailrosen]

Hi, It doesn't really, because all I'm trying to do is run the Virtual Petri dish example, and I notice that I am setting all of that and more with the example you have (so not sure why it doesn't work): p.all_params["maxCycles"] = 5000 p.all_params["biomassMotionStyle"] = "Convection 2D" p.all_params["writeBiomassLog"] = True p.all_params["BiomassLogRate"] = p.all_params["maxCycles"] p.all_params["defaultKm"] = 0.01 p.all_params["defaultVmax"] = 10 p.all_params["timeStep"] = 0.01 I try to run the simulation on Windows.. and it still can't find Gurobi even though I have set it manually to a hundred places/variables -- see the attached ipynb. p.all_params["spaceWidth"] = 0.02 p.all_params["maxSpaceBiomass"] = 10 p.all_params["minSpaceBiomass"] = 1.e-10 p.all_params["allowCellOverlap"] = True p.all_params["writeFluxLog"] = True p.all_params["FluxLogRate"] = p.all_params["maxCycles"] p.all_params["writeMediaLog"] = True p.all_params["MediaLogRate"] = p.all_params["maxCycles"] And when I go back to windows, it still complains that JAVA cannot find Gurobi.. even though I set the path in a variety of ways (see attached). thanks, Gail

…

________________________________________ From: dukovski ***@***.***> Sent: Thursday, January 23, 2025 9:53 AM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Yes, you need to set the parameter to save the biomass. The total_biomass is saved by default, but that is the integrated biomass. If you want to save the spatial distribution of the biomass you need to specify: p = cometspy.params() p.set_param("writeBiomassLog", True) also set the rate of writing the log. I hope this helps. — Reply to this email directly, view it on GitHub<#54 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABHZROP4R6VNV7E7KPRP4VL2MD65RAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMJQGAZDANJZGA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[gailrosen]

btw.. this is the mac python code

…

________________________________________ From: Rosen,Gail ***@***.***> Sent: Thursday, January 23, 2025 12:25 PM To: segrelab/comets Subject: Re: [segrelab/comets] Container? (Issue #54) Hi, It doesn't really, because all I'm trying to do is run the Virtual Petri dish example, and I notice that I am setting all of that and more with the example you have (so not sure why it doesn't work): p.all_params["maxCycles"] = 5000 p.all_params["biomassMotionStyle"] = "Convection 2D" p.all_params["writeBiomassLog"] = True p.all_params["BiomassLogRate"] = p.all_params["maxCycles"] p.all_params["defaultKm"] = 0.01 p.all_params["defaultVmax"] = 10 p.all_params["timeStep"] = 0.01 I try to run the simulation on Windows.. and it still can't find Gurobi even though I have set it manually to a hundred places/variables -- see the attached ipynb. p.all_params["spaceWidth"] = 0.02 p.all_params["maxSpaceBiomass"] = 10 p.all_params["minSpaceBiomass"] = 1.e-10 p.all_params["allowCellOverlap"] = True p.all_params["writeFluxLog"] = True p.all_params["FluxLogRate"] = p.all_params["maxCycles"] p.all_params["writeMediaLog"] = True p.all_params["MediaLogRate"] = p.all_params["maxCycles"] And when I go back to windows, it still complains that JAVA cannot find Gurobi.. even though I set the path in a variety of ways (see attached). thanks, Gail

________________________________________ From: dukovski ***@***.***> Sent: Thursday, January 23, 2025 9:53 AM To: segrelab/comets Cc: Rosen,Gail; Author Subject: Re: [segrelab/comets] Container? (Issue #54) External. Yes, you need to set the parameter to save the biomass. The total_biomass is saved by default, but that is the integrated biomass. If you want to save the spatial distribution of the biomass you need to specify: p = cometspy.params() p.set_param("writeBiomassLog", True) also set the rate of writing the log. I hope this helps. — Reply to this email directly, view it on GitHub<#54 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ABHZROP4R6VNV7E7KPRP4VL2MD65RAVCNFSM6AAAAABVK4IS2SVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMJQGAZDANJZGA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[gailrosen]

can you see my attachments on email?