mbakker7 / ttim

MIT License
33 stars 23 forks source link

Circular Inhomogeneity results in IndexError #2

Closed Huite closed 9 months ago

Huite commented 6 years ago

(Zoals ik donderdag al even meldde:) Adding a circular inhomogeneity leads to an IndexError. The following (simplified) case runs fine:

import ttim

ml = ttim.ModelMaq(kaq=[30],z=[0,-10,-50],Saq=[0.001],
                   c=[1500], Sll=[0.01],
                   topboundary='leaky', phreatictop=True,tmin=1,tmax=30,M=20)

ttim.Well(ml,xw=0,yw=0,rw=0.5,tsandQ=[(4.0,2500),(25,0.),(30,0.)],
res=0.0,layers=0,rc=None,wbstype='pumping',label=None)

ml.solve()

But adding:

ttim.CircInhomMaq(ml,x0=0,y0=0,R=5,order=1,
                  kaq=[30],z=[0,-10,-50],Saq=[0.001],
                   c=[1500], Sll=[0.01],
                   topboundary='leaky', phreatictop=True)
ml.solve()

Results in the following error:

Traceback (most recent call last):

  File "<ipython-input-7-35237950a7fc>", line 14, in <module>
    ml.solve()

  File "D:\Anaconda\envs\py27\lib\site-packages\ttim\ttim.py", line 277, in solve
    mat[ ieq:ieq+e.Nunknowns, :, : ], rhs[ ieq:ieq+e.Nunknowns, :, : ] = e.equation()

  File "D:\Anaconda\envs\py27\lib\site-packages\ttim\ttim.py", line 1088, in equation
    e.potinflayers(self.xc[icp],self.yc[icp],self.pylayers,self.aqin) / self.aqin.T[self.pylayers][:,np.newaxis,np.newaxis] - \

  File "D:\Anaconda\envs\py27\lib\site-packages\ttim\ttim.py", line 670, in potinflayers
    return rv[pylayers,:]

IndexError: index 1 is out of bounds for axis 0 with size 1

Using more layers results in the same error message ("Index N is out of bounds for axis 0 with size N").

As an aside, with regards to zero indexing in general, it looks like ModelMaq() does not accept tmin=0. (Of course, this is not a serious hindrance.)

Met vriendelijke groeten, Huite

mjigmond commented 6 years ago

See the following commit 44145c0. You can checkout the master branch or do a manual fix on your machine.

dbrakenhoff commented 9 months ago

I'm guessing this one is fixed, given the referenced commit in the previous post. Reopen if that's not the case.