I would like to build something like this if statements into a numpy condition (for argwhere or similar).
for i, element in enumerate(array):
if array[i]+1 == array[i+1]:
return True
Is this possible or is looping through the array the only way?
You can use the np.roll function that shifts the array
import numpy as np
array+1 == np.roll(array, shift=-1)
This will return an array of boolean values.
shift = -1
You can use np.equal() function as well.
import numpy as np
a = [1,2,4,8,12,13]
x = np.array(a)
res = np.equal((x+1)[:-1], x[1:])
print(res)
Out:
[ True False False False True]
Note:
If you need speed, as it is usual the case, when we use numpy, it is worth to mention, that this methode is faster then np.roll(), what FBruzzesi proposed below and which is an elegant solution as well anyway:
import timeit
code1 = """
import numpy as np
a = [1,2,4,8,12,13]
x = np.array(a)
np.equal((x+1)[:-1], x[1:])
"""
elapsed_time1 = timeit.timeit(code1, number=10000)/100
print(elapsed_time1)
code2 = """
import numpy as np
a = [1,2,4,8,12,13]
x = np.array(a)
x+1 == np.roll(x, shift=-1)
"""
elapsed_time2 = timeit.timeit(code2, number=10000)/100
print(elapsed_time2)
Out:
0.00044608700000026147
0.0022752689999970244
