Dictionary not recognizing floating point keys

问题:

I have a dictionary called G. When I enter G. keys (), an example of the output is:

>>> G.keys ()
[(1490775.0, 12037425.0), (1493775.0, 12042675.0), (1481055.0, 12046305.0), (1503105.0, 12047415.0), (1488585.0, 12050685.0), (1483935.0, 12051405.0),...

When I use the operation key in G the result is false.

>>> (1490775.0, 12037425.0) in G
False

Why isn't my dictionary recognizing my keys?

>>> type (G.keys()[0])
<type 'numpy.void'>
>>> type (G.keys()[0][0])
<type 'numpy.float64'>
>>> type (G.keys()[0][1])
<type 'numpy.float64'>
type(G)
<type 'dict'>

回答1:

This might be how you arrived in this situation:

import numpy as np
arr = np.array([(1490775.0, 12037425.0)], dtype=[('foo','<f8'),('bar','<f8')])
arr.flags.writeable = False

G = dict()
G[arr[0]] = 0

print(type(G.keys()[0]))
# <type 'numpy.void'>

print(type(G.keys()[0][0]))
# <type 'numpy.float64'>

print(type(G.keys()[0][1]))
# <type 'numpy.float64'>

print(type(G))
# <type 'dict'>

A tuple of floats is not a key in G:

print((1490775.0, 12037425.0) in G)
# False

But the numpy.void instance is a key in G:

print(arr[0] in G)
# True

---

You will probaby be better off not using numpy.voids as keys. Instead, if you really need a dict, then perhaps convert the array to a list first:

In [173]: arr.tolist()
Out[173]: [(1490775.0, 12037425.0)]
In [174]: G = {item:0 for item in arr.tolist()}

In [175]: G
Out[175]: {(1490775.0, 12037425.0): 0}

In [176]: (1490775.0, 12037425.0) in G
Out[176]: True