Issue with Multi-Simulation Error during Data Label Generation

[ng-fukgin]

Hello,

I've encountered an issue while trying to generate labels with the following command:

python main.py --simulate --env chaohu --data_dir edge_hc --edge_fusion

The error message received was:

~/surrogate$ CUDA_VISIBLE_DEVICES=1 python main.py --simulate --env chaohu --data_dir edge_hc --edge_fusion 
Training configs: Namespace(act='False', activation='relu', balance=0, batch_size=256, conv='GCNconv', data_dir='./envs/data/chaohu/edge_hc', directed=False, dropout=0.0, edge_fusion=True, embed_size=128, env='chaohu', epochs=500, epsilon=-1.0, hidden_dim=64, hotstart=False, if_flood=0, kernel_size=3, learning_rate=0.001, length=0, load_model=False, model_dir='./model/chaohu/', n_out=3, n_sp_layer=2, n_tp_layer=2, norm=False, order=1, processes=5, rain_dir='./envs/config/', rain_num=50, rain_suffix='chaohu', ratio=0.8, recurrent='GRU', repeats=1, resnet=False, result_dir='./results/chaohu/', save_gap=500, seq_in=6, seq_out=1, setting_duration=5, simulate=True, test=False, train=False, use_adj=False, use_edge=False)
multiprocessing.pool.RemoteTraceback: 
"""
Traceback (most recent call last):
  File "/raid/home/dgxuser4/anaconda3/envs/GNN-UDS/lib/python3.8/multiprocessing/pool.py", line 125, in worker
    result = (True, func(*args, **kwds))
  File "/raid/home/dgxuser4/raid/hydraulic_prediction/GNN-UDS/surrogate/dataloader.py", line 30, in simulate
    state = self.env.reset(event,global_state=True,seq=seq)
  File "/raid/home/dgxuser4/raid/hydraulic_prediction/GNN-UDS/surrogate/envs/scenario/base.py", line 208, in reset
    self.env = env_base(self.config, ctrl=True)
  File "/raid/home/dgxuser4/raid/hydraulic_prediction/GNN-UDS/surrogate/envs/environment/environment_base.py", line 50, in __init__
    super().__init__(config, ctrl, binary)
  File "/raid/home/dgxuser4/anaconda3/envs/GNN-UDS/lib/python3.8/site-packages/pystorms/environment.py", line 52, in __init__
    self.sim = Simulation(self.config["swmm_input"])
  File "/raid/home/dgxuser4/anaconda3/envs/GNN-UDS/lib/python3.8/site-packages/pyswmm/simulation.py", line 100, in __init__
    raise(MultiSimulationError("Multi-Simulation Error."))
pyswmm.errors.MultiSimulationError: Multi-Simulation Error.

        Due to architectural limitations inside of EPA-SWMM
        multiple Simulations cannot be completed within a single
        instance of Python.  Consider launching the model runs
        using a Python Subprocess and a Queue manager.

"""

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "main.py", line 158, in <module>
    dG.generate(events,processes=args.processes,repeats=args.repeats,act=args.act)
  File "/raid/home/dgxuser4/raid/hydraulic_prediction/GNN-UDS/surrogate/dataloader.py", line 71, in generate
    res = [r.get() for r in res]
  File "/raid/home/dgxuser4/raid/hydraulic_prediction/GNN-UDS/surrogate/dataloader.py", line 71, in <listcomp>
    res = [r.get() for r in res]
  File "/raid/home/dgxuser4/anaconda3/envs/GNN-UDS/lib/python3.8/multiprocessing/pool.py", line 768, in get
    raise self._value
pyswmm.errors.MultiSimulationError: Multi-Simulation Error.

        Due to architectural limitations inside of EPA-SWMM
        multiple Simulations cannot be completed within a single
        instance of Python.  Consider launching the model runs
        using a Python Subprocess and a Queue manager.


        Due to architectural limitations inside of EPA-SWMM
        multiple Simulations cannot be completed within a single
        instance of Python.  Consider launching the model runs
        using a Python Subprocess and a Queue manager.

This suggests that the SWMM simulations cannot be executed concurrently in a single Python instance. I am running this on an Ubuntu system with Python 3.8.2. I followed the steps outlined in the README for generating data. Could you please advise on how to resolve this or provide any additional steps I may need to ensure proper execution?

Thank you!

[ng-fukgin]

Hello,

I've managed to resolve the issue previously reported. Initially, I had installed libraries from requirements.txt, which led to the problem. I have now installed the libraries from requirements2.txt, which has fixed the issue.

Thank you for the support!

[Zhiyu014]

Thanks for your interests in this repo. This issue comes from the pyswmm 2.0 update, as I'm still using pyswmm==1.5.1.