Setting a relative model path fails when creating the first model

[Hugovdberg]

When I run the following snippet I get a TypeError:

import flopy
sim = flopy.mf6.MFSimulation(sim_name="sim", sim_ws='models')
tdis = flopy.mf6.ModflowTdis(sim, time_units="DAYS")
gwf = flopy.mf6.ModflowGwf(sim, modelname='base', model_rel_path='base')

Details

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-1-3901b9f6c6b6> in <module>
      2 sim = flopy.mf6.MFSimulation(sim_name="sim", sim_ws='models')
      3 tdis = flopy.mf6.ModflowTdis(sim, time_units="DAYS")
----> 4 gwf = flopy.mf6.ModflowGwf(sim, modelname='base', model_rel_path='base')

C:\Anaconda64\lib\site-packages\flopy\mf6\modflow\mfgwf.py in __init__(self, simulation, modelname, model_nam_file, version, exe_name, model_rel_path, list, print_input, print_flows, save_flows, newtonoptions, packages, **kwargs)
     85                                          version=version, exe_name=exe_name,
     86                                          model_rel_path=model_rel_path,
---> 87                                          **kwargs)
     88 
     89         self.name_file.list.set_data(list)

C:\Anaconda64\lib\site-packages\flopy\mf6\mfmodel.py in __init__(self, simulation, model_type, modelname, model_nam_file, version, exe_name, add_to_simulation, structure, model_rel_path, verbose, **kwargs)
    102             model_nam_file = '{}.nam'.format(modelname)
    103 
--> 104         self.set_model_relative_path(model_rel_path)
    105         if add_to_simulation:
    106             self.structure = simulation.register_model(self, model_type,

C:\Anaconda64\lib\site-packages\flopy\mf6\mfmodel.py in set_model_relative_path(self, model_ws)
    700             models = self.simulation.name_file.models
    701             models_data = models.get_data()
--> 702             for index, entry in enumerate(models_data):
    703                 old_model_file_name = os.path.split(entry[1])[1]
    704                 old_model_base_name = os.path.splitext(old_model_file_name)[0]

TypeError: 'NoneType' object is not iterable

It appears to fail because I set a relative path immediately for the first model, when the simulation does not contain any models with a path yet. The workaround to use model_rel_path='.' and call set_model_relative_path manually aftewards also fails:

sim = flopy.mf6.MFSimulation(sim_name="sim", sim_ws='models')
tdis = flopy.mf6.ModflowTdis(sim, time_units="DAYS")
gwf = flopy.mf6.ModflowGwf(sim, modelname='base')
gwf.set_model_relative_path('base')

Details

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-3-36551b578161> in <module>
      2 tdis = flopy.mf6.ModflowTdis(sim, time_units="DAYS")
      3 gwf = flopy.mf6.ModflowGwf(sim, modelname='base')
----> 4 gwf.set_model_relative_path('base')

C:\Anaconda64\lib\site-packages\flopy\mf6\mfmodel.py in set_model_relative_path(self, model_ws)
    732                 packages = self.name_file.packages
    733                 packages_data = packages.get_data()
--> 734                 for index, entry in enumerate(packages_data):
    735                     old_package_name = os.path.split(entry[1])[1]
    736                     packages_data[index][1] = os.path.join(path,

TypeError: 'NoneType' object is not iterable

This now seems to be caused by the fact that the model doesn't contain any packages yet, confirmed by the fact the following code does work as it's supposed:

sim = flopy.mf6.MFSimulation(sim_name="sim", sim_ws='models')
tdis = flopy.mf6.ModflowTdis(sim, time_units="DAYS")
gwf = flopy.mf6.ModflowGwf(sim, modelname='base')
dis = flopy.mf6.ModflowGwfdis(gwf)
gwf.set_model_relative_path('base')

[spaulins-usgs]

I checked in a fix that should solve the problem. Please try checking out the latest flopy code on the develop branch and let me know if it fixes the problem for you.

[Hugovdberg]

Sorry for the late reply, it appears this did fix the problem, thanks a lot!