a Python toolbox for neural data analysis, developed in Sheinberg lab at Brown University
by Shaobo Guan and Ruobing Xia, created in 2016
This package provides a universual data structure that meets most neurophysiologicla data analysis demands, and provides various functions for bacis and advanced data analsis and visualiztion, including rasters, PSTH, ERP, tuning curve, population decoding, dimensionality reduction, information measurment, variance and correlation analsysis, spectrum analysis, functinal connectivity, current source density (CSD) analsis, recording profile analsis etc.
- typical python moduels:
- numpy
- scipy
- matplotlib
- pandas
- sklearn
- other modules:
- dgread (optional for reading stimdg behavioral data, could use pandas instead if behaviral data is in pandas DataFrame)
- python-neo
- dependencis of neo
- Neurophysiology data: any widely used data format in neurophysiology as long as it is supported by NEO object IO, e.g. TDT, Plexon, BlackRock, and etc.
- Behavioral data: pandas Dataframe or DLSH dynamic group (dg)
The core data is a 3D array of shape [number_of_trials, number_of_timestamps, number_of_channels] that works for both spiking data and LFP/EEG data, plus a pandas DataFrame that stored task/behaviorial information related to every trial
- A getting started guide can be found at
/demo_script/Tutorial_0_Basic_Data_Structure_data_neuro.ipynb - A small proportion of implemented functionalities is demonstrated in tutorials, which can be found in
/demo_script - More usages can be found in the folder
day_noteandscript - For detailed usage of a particular function, go to the source code or the documentation page, or just check the doc string by typing
help(module_or_function_of_interest)