ndarraybuffer 0.2.1

NDArray with both-side insertion/deletion and size limit

Introduction

Wrap numpy.ndarray to support fast both-side insertion/deletion and size limit.

• Fully typed.
• Directly perform arithmetic operations with NDArray
• pickleable
• JSON serializable

pip install ndarraybuffer

Both-side insertion/deletion

arr = ArrayBuffer()
arr.extend(np.arange(100))
arr.append(100)
arr.appendleft(0)
arr.pop(1)
arr.popleft(1)

Size limit

arr = ArrayBuffer(max_len=10)
arr.extend(np.arange(10))
assert len(arr) == 10
assert np.array_equal(arr, np.arange(10))
arr.append(10)
assert len(arr) == 10
assert np.array_equal(arr, np.arange(1, 11))

JSON Serialization and Deserialization

arr = ArrayBuffer(max_len=10, dtype=np.float64)
arr.extend(np.arange(10))
state_dict = arr.state_dict()
rec = ArrayBuffer.load(json.loads(json.dumps(state_dict)))