I'm trying to create a dictionary that maps a specific array like [1,1,0,0] to a string 'Car' but the dictionary does not accept arrays or lists
a={(1,1,0,0):'Car',
(0,0,0,1):'Pedestrian',
(1,0,0,0):'Traffic Light'}
b=np.array([[1,1,0,0],[1,0,0,0],[0,0,0,1]])
Both codes are free of errors but obviously they don't match. Here is my idea:
b.map(a)
Out[3]=['Car','Traffic Light','Pedestrian']
Thanks in advance.
A simple list comprehension will do, converting each inner array to a tuple before passing it to the dict for lookup:
>>> out = [a[tuple(x)] for x in b]
>>> out
['Car', 'Traffic Light', 'Pedestrian']
You can grab the string using a tuple as a key. The reason you can't use a list as a key is because lists can be modified, and you wouldn't other spots in your code modifying your dictionary keys.
>>> a[[1,1,0,0]] # Won't work because it's not a tuple
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: unhashable type: 'list'
>>> a[(1,1,0,0)] # Works
'Car'
If you want to use the list as a key, then you can convert it
>>> a[tuple([1,1,0,0])]
'Car'
If you have a list of arrays (like in your example) then you can grab all their matching strings like this
>>> b=np.array([[1,1,0,0],[1,0,0,0],[0,0,0,1]])
>>> [a[tuple(eachList)] for eachList in b]
['Car','Traffic Light','Pedestrian']
If you are looking after efficiency if b is a big array, transform it in hashable keys :
B=(b<<np.arange(4)).sum(1)
# array([3, 1, 8], dtype=int32)
Then define dict as :
A={1: 'Traffic Light', 3: 'Car', 8: 'Pedestrian'}
The mapping now can be:
[A[key] for key in B]
dictobjects cannot havelistornp.arrayobjects as keys. Perhapstuple(my_list)as a key?mapas defined. Where are you stuck?np.ndarrayobjects don't have a.mapmethod...