Fill a numpy array with objects

Question

I have code that sometimes raises an AttributeError. The exact error is :AttributeError: 'int' object has no attribute 'place' I create a numpy array with a=np.eye(9, dtype=object) then fill it with the objects as follows:

for i in range(9): #rows
            #Create the initial Place for each row in the column 0
            j=0
            #Check if there is something already stored if not store a new object
            try:
                places[i,j].place #An attribute of the 
            except:
                places[i,j]=Place() #Where Place() is the object (it has the attribute place)

            for j in range(9): #columns
            #For the other columns of the first row(i==0) create a Place according to the previous Place
                if i==0 and j>=1:
                    #If there is something already don't create a new Place
                    try: 
                        places[i,j].place
                    except AttributeError:
                        if places[i,j-1].place=='cave':
                            places[i,j]=Place()
                if i>=1 and j>=1:
                    try:
                        places[i,j].place
                    except:
                        print(places)
                        if places[i,j-1].place=='cave' and places[i-1,j].place=='cave':
                            places[i,j]=Place()

When I run this code it usually stops at the last if: if places[i,j-1].place=='cave' and places[i-1,j].place=='cave':. But it should work fine, shouldn't it?

I put the print(places) in to check how this process runs and how it is filled. Here is the result for the first run (it has some other object of previous conditions not relevant with the problem, or maybe it is):

[[ 
   
   
   
  ]
 [ 
   0 0 0 0 0 0]
 [0 
   0 0 0 0 0 0]
 [0 0  1 0 0 0 0 0]
 [0 0 0 0 1 0 0 0 0]
 [0 0 0 0 0 
   0 0]
 [0 0 0 0 0 0 1  0]
 [0 0 0 0 0 0 0 
  ]
 [0 0 0 0 0 0 0 
  ]]

Here is the last one before the error where apparently everything has worked, so the code seems to work fine:

But even now, it throws the error of AttributeError: 'int' object has no attribute 'place'. As seen above just the last row needs to be filled but in the previous iteration it behaves as if the conditions to fill the map weren't enough, when in the previous lines they were.

Any idea why this is happening, and what can I do to fix it?

user2357112 · Accepted Answer

                try: 
                    places[i,j].place
                except AttributeError:
                    if places[i,j-1].place=='cave':
                        places[i,j]=Place()

Do all entries of your array need to be Place objects? For columns after the first, you only sometimes fill them with Places. If places[i, j-1].place wasn't 'cave', you'll never fill places[i, j], and then on the next iteration, places[i, j-1].place will raise an AttributeError. This happens on the if i>=1 and j>=1: branch, too.

Why not fill the array with Places at creation time, so you don't have to keep checking whether it has Places in it?

places = numpy.array([[Place() for i in xrange(n)] for j in xrange(n)])

Fill a numpy array with objects

Answers (1)

Related Questions