numpy_array_buffer

Library for buffers used in real-time computation:

• RingBuffer: when full overwrites oldest row
• DownsamplingBuffer: when full downsamples full buffer

Example

ring_buffer = RingBuffer(3,2, column_names=["x", "y"])
for i in range(6):
    ring_buffer.append([i, i + 10])
    print(f"{i=}, y: {ring_buffer.get_col('x')}")

output:

i=0, y: [0.]
i=1, y: [0. 1.]
i=2, y: [0. 1. 2.]
i=3, y: [1. 2. 3.]
i=4, y: [2. 3. 4.]
i=5, y: [3. 4. 5.]

Downsampled array

downsampling_buffer = DownsamplingBuffer(4, 2)
for i in range(9):
    data_object = [i, i + 10]
    downsampling_buffer.append(data_object)
    print(f"{i=}, Arr: {downsampling_buffer.get_array()[:,0]}")

output:

i=0, Arr: [0.]
i=1, Arr: [0. 1.]
i=2, Arr: [0. 1. 2.]
i=3, Arr: [0. 2.]
i=4, Arr: [0. 2. 4.]
i=5, Arr: [0. 4.]
i=6, Arr: [0. 4.]
i=7, Arr: [0. 4.]
i=8, Arr: [0. 4. 8.]

Similar packages:

Look at if functionality is the same, or compare benchmark:

• https://github.com/eric-wieser/numpy_ringbuffer/blob/master/numpy_ringbuffer/__init__.py
• https://stackoverflow.com/questions/8908998/ring-buffer-with-numpy-ctypes

Develop:

pip install -e .