CircularBuffer

class torch_sim.properties.correlations.CircularBuffer(size, device=None)

Bases: object

Circular buffer for storing time series data.

Provides a fixed-size circular buffer optimized for storing and retrieving time series data, with minimal memory allocation.

Variables:

• size – Maximum number of elements to store

• buffer – Storage for the data

• head – Current write position

• count – Number of elements currently stored

• device – Device where the buffer is stored

Parameters:

• size (int)

• device (device | None)

append(value)

Append a new value to the buffer.

Parameters:

value (Tensor) – New tensor to store

Return type:

None

get_array()

Get the current buffer contents as a tensor.

Returns:

Tensor containing the buffered data in chron. order

Return type:

Tensor

property is_full: bool

Check if the buffer is full.