Ring (aka circular) buffers backed by a numpy array, supporting:

• Operations from collections.deque
  • b.append(val)
  • b.appendleft(val)
  • b.extend(val)
  • b.extendleft(val)
  • b.pop(val)
  • b.popleft(val)
• The collections.Sequence protocol (unoptimized)
• C-side unwrapping into an array with np.array(b)
• Arbitrary element dtypes, including extra dimensions like RingBuffer(N, dtype=(int, 3))

For example:

import numpy as np from numpy_ringbuffer import RingBuffer r = RingBuffer(capacity=4, dtype=np.bool) r.append(True) r.appendleft(False) print(np.array(r)) # array([False, True])