smlmlp.metric_nena module

smlmlp.metric_nena(col, blk, fr, *, bins=101, cuda=False, parallel=False)[source]

Estimate one-dimensional NeNA precision from consecutive points in each blink.

Parameters:

  • col (array-like) – Scalar localization values used for NeNA differences.

  • blk (array-like) – Blink index for each localization.

  • fr (array-like) – Frame identifiers used to order localizations inside each blink.

  • bins (int or array-like, optional) – Histogram bin specification passed to numpy.histogram().

  • cuda (bool or int, optional) – Execution options accepted by all analysis functions. cuda is kept for API consistency and is not used in this CPU implementation.

  • parallel (bool or int, optional) – Execution options accepted by all analysis functions. cuda is kept for API consistency and is not used in this CPU implementation.

Returns:

  • precision (float) – One-dimensional NeNA precision, computed as sigma / sqrt(2) from the fitted Gaussian width.

  • info (dict) – Diagnostic information with differences, histogram, fit parameters, and blink counts.