–Compression Ratio
–Size (Original → Compressed)
–Max Reconstruction Error
–Execution Time
The error bound is a percentage of the data range (max−min): it caps the maximum per-point deviation for most methods, bounds the RMSE for Bottom-up and Sliding-window, and the area error for Visvalingam–Whyatt. DFT is swept by its number of retained coefficients, and BitPackedBUFF by its decimal precision (digits kept).
Build new compression methods by merging existing ones into a custom pipeline. Coming soon.
Search for the method and error bound that best balance compression ratio and reconstruction error on your series. Coming soon.