CODEWITHSUNDEEP
python
Jackie
Jackie

Reputation: 307

python - Efficient way to find the nearest point in 2-Dimensional

I have 100 stations and point A(x,y)
I need to find the nearest station to A. I found the solution here, but it throws an error.

Here is what I did:

def closest_node(node, listStation):
   nodes = []
   for i in xrange(1,len(listStation)):
       nodes.append((listStation[i][12],listStation[i][13]))
   nodes = np.asarray(nodes)
   dist_2 = np.sum((nodes - node)**2, axis=1)
   return np.argmin(dist_2)

print closest_node(((99,100)), listStation)

And here is the error:

dist_2 = np.sum((nodes - node)**2, axis=1)
TypeError: unsupported operand type(s) for -: 'numpy.ndarray' and 'tuple'

Thanks to @WeaselFox, the difference is I got the data from CSV file. here is the solution:

def closest_node(node, listStation):
  nodes = []
  for i in xrange(1,len(listStation)):
      a = np.asanyarray((float(listStation[i][12]),float(listStation[i][13])))
      nodes.append(a)        

  nodes = np.asarray(nodes)    
  dist_2 = np.sum((nodes - node)**2, axis=1)
  return np.argmin(dist_2)

Upvotes: 0

Views: 202

Answers (1)

WeaselFox
WeaselFox

Reputation: 7380

you need to cast your node to an np array for numpy to calculate the distance:

node = np.asarray(node)

example:

>>> a = np.asarray((1,2))
>>> b = np.asarray((2,3))
>>> c = np.asarray((1,1))
>>> nodes = [b,c]
>>> dist_2 = np.sum((a - nodes)**2, axis=1)
>>> nodes[np.argmin(dist_2)]
array([1, 1])

Upvotes: 1

Related Questions