Axis out of bounds

[jsevo]

Hi there,

I tried using your code on my own data and encountered several issues. It appears to boil down to an axis out of bounds issue.

Running your sample code (using python 2.7.3 on linux) returns the following error:
"""
d, i = T.query([45.0, 10.0, 10.0], k=4)
File "/usr/local/lib/python2.7/dist-packages/periodic_kdtree.py", line 306, in query
hits = self.__query(x, k=k, eps=eps, p=p, distance_upper_bound=distance_upper_bound)
File "/usr/local/lib/python2.7/dist-packages/periodic_kdtree.py", line 217, in __query
self.max_distance_upper_bound)
File "/usr/lib/python2.7/dist-packages/numpy/core/fromnumeric.py", line 1894, in amin
return _wrapit(a, 'min', axis, out)
File "/usr/lib/python2.7/dist-packages/numpy/core/fromnumeric.py", line 37, in _wrapit
result = getattr(asarray(obj),method)(_args, *_kwds)
ValueError: axis(=15) out of bounds

"""

[jsevo]

I found the bit that caused the bug but I am unsure how to proceed:

Line 216-217 of periodic_kdtree.py

It appears that the following (class PeriodiccKDTree(KDTree)):

distance_upper_bound = np.min(distance_upper_bound, self.max_distance_upper_bound)

needs to be changed to (note the square brackets)

distance_upper_bound = np.min([distance_upper_bound, self.max_distance_upper_bound])

EDIT:

The same applies for Line 131 (class PeriodicKDTree(KDTree)):

r = np.min(r, self.max_distance_upper_bound)

has to be changed to

r = np.min([r, self.max_distance_upper_bound])

[pdebuyl]

Hi,

I have the exact same issue. From np.min docstring, it is clear that the second argument is axis and not an item to be compared.

Using NumPy 1.7.1

I'd be happy to make this a pull request if @patvarilly is ok with submissions.