Numpy find indices of groups with same value

We can do something like this that works for any generic array –

def islandinfo(y, trigger_val, stopind_inclusive=True):
    # Setup "sentients" on either sides to make sure we have setup
    # "ramps" to catch the start and stop for the edge islands
    # (left-most and right-most islands) respectively
    y_ext = np.r_[False,y==trigger_val, False]

    # Get indices of shifts, which represent the start and stop indices
    idx = np.flatnonzero(y_ext[:-1] != y_ext[1:])

    # Lengths of islands if needed
    lens = idx[1::2] - idx[:-1:2]

    # Using a stepsize of 2 would get us start and stop indices for each island
    return list(zip(idx[:-1:2], idx[1::2]-int(stopind_inclusive))), lens

Sample run –

In [320]: y
Out[320]: array([1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1])

In [321]: islandinfo(y, trigger_val=1)[0]
Out[321]: [(0, 2), (8, 9), (16, 19)]

In [322]: islandinfo(y, trigger_val=0)[0]
Out[322]: [(3, 7), (10, 15)]

Alternatively, we can use diff to get the sliced comparisons and then simply reshape with 2 columns to replace the step-sized slicing to give ourselves a one-liner –

In [300]: np.flatnonzero(np.diff(np.r_[0,y,0])!=0).reshape(-1,2) - [0,1]
Out[300]: 
array([[ 0,  2],
       [ 8,  9],
       [16, 19]])