numpy.searchsorted(a, v, side='left', sorter=None) → ndarray
Supports 1D.
cupy.searchsorted(a, v, side='left', sorter=None) → ndarray
Supports 1D.
dask.array.searchsorted(a, v, side='left', sorter=None) → ndarray
jax.numpy.searchsorted(a, v, side='left', sorter=None, *, method='scan') → ndarray
method controls device and array size performance optimizations.
np.searchsorted(a, v, side='left', sorter=None) → ndarray
torch.searchsorted(sorted_sequence, values, *, out_int32=False, right=False, side='left', out=None, sorter=None) → Tensor
side is preferred over right. Supports N-D (supports stacking).
tf.searchsorted(sorted_sequence, values, side='left', out_type=tf.dtypes.int32, name=None) → Tensor
Does not support sorter.