Replay-classification

Latest version: v0.6.1

Safety actively analyzes 638841 Python packages for vulnerabilities to keep your Python projects secure.

Scan your dependencies

Page 2 of 2

0.3.2

+ Implements a L2-regularized GLM to fit sorted spike model so that convergence issues are avoided
+ Implements a scaled likelihood for numerical stability (computations are done in log likelihood space then converted to a scaled likelihood where the max value is 1).
+ Handle NaN in trajectory direction in sorted spike decoder
+ Count first crossing as the predicted state for the sorted spike decoder to match the Clusterless decoder

0.3.1

This release updates how the code decides the replay state. Previously, the code found the state that had the highest confidence level at the last time point. Now, the code determines the state by the first state to cross a pre-defined confidence level which can be set by the `confidence_threshold` parameter when initializing the class.

0.3.0

+ Fix conditioning of empirical movement state transition matrix (04438d974dbd2646304569973a9ef48af19d7813)
+ Make default place std deviation larger to do more smoothing (099ecc17dac0119b8520e7bf730d35946440cb7c)
+ Add option to specify place standard deviation (9238f0b6470c0d46decd518bec9f54dc72703ada)
+ Fix array shape of spikes in sorted spikes decoder to match clusterless decoder (4ea36555c9e7876ee8d03b87d9e19a1260ec15d2)
+ Add module for simulating marked point process and point process place fields (87dbc9e4a0393ecfae635b7c6b373e7d70b0a9f0)
+ Update clusterless decoder notebook with simulation

0.2.1

+ Fixed problem with state transition matrix where columns summed to one, not rows.
+ Decoding results now return posterior, likelihood, and prior for debugging purposes
+ New interactive plot for simultaneously visualizing posterior, likelihood, and prior at each time point for debugging purposes
+ Removed unused debug keyword argument in `predict state`

0.1.1

Handle NaNs in trajectory direction

0.1.0

This initial release implements code from:

> Deng, X., Liu, D.F., Karlsson, M.P., Frank, L.M., and Eden, U.T. (2016). Rapid classification of hippocampal replay content for real-time applications. Journal of Neurophysiology 116, 2221–2235.

Page 2 of 2

© 2024 Safety CLI Cybersecurity Inc. All Rights Reserved.