I have an object -> {0: 0.8, 1: 0.2, 2: 0, 3: 0}
I want to convert it to a numpy array, where the keys are the index, and the values are the values of the array -> [0.8, 0.2, 0, 0]
Whats the fastest and efficient way of doing this?
I am using a for loop, but is there a better way of doing this?
Assuming your dictionary is called dict:
numpy_array = np.array([*dict.values()])
keys, items and values are the fastest way to get 'things' from a dict. Otherwise you have to iterate on the keys. A general approach that allows for skipped indices, and out-of-order ones:
In [81]: adict = {0: 0.8, 1: 0.2, 2: 0, 3: 0}
In [82]: keys = list(adict.keys())
In [83]: arr = np.zeros(max(keys)+1) # or set your own size
In [84]: arr[keys] = list(adict.values())
In [85]: arr
Out[85]: array([0.8, 0.2, 0. , 0. ])
The above answer gives dict_values, but was in the right direction
the correct way to do it is:
d = {0: 0.8, 1: 0.2, 2: 0, 3: 0}
np_array = np.array(list(d.values()))
Dictionary is inherently orderless. You would need to sort your values according to your keys (assuming you do not have a missing key in your dictionary):
a = {0: 0.8, 1: 0.2, 2: 0, 3: 0}
np.array([i[1] for i in sorted(a.items(), key=lambda x:x[0])])
Another way of doing it is sorting in numpy:
b = np.array(list(a.items()))
b[b[:,0].argsort()][:,1]
output:
[0.8, 0.2, 0, 0]
A solution if input dict is not ordered or has missing values:
d = {1:1.1, 2:2.2, 0: 0.8, 4:4.4}
sz = 1+max(d.keys()) # or len(d) if you are sure there are no missing values
x = np.full(sz, np.nan)
x[list(d.keys())] = list(d.values())
x
#array([0.8, 1.1, 2.2, nan, 4.4])
d.values()answers below are correct for your ordered input (assuming you have py >3.6 where dict (stackoverflow.com/questions/39980323/…) gained insertion order). But is there a chance that your dict is input in a "wrong" order? For exampled = {1:1.1, 2:2.2, 0: 0.8}