* add Nest backend and model class
* adding nest backend
* add joint_test class
* add/remove blank lines to comply with PEP8
* fix typo; simulaton --> simulation
* add self to __ini__ function arguments
* update call to self.set_default_run_params
* remove whitespaces to comply with PEP8
* adding backends to setup.py
* backend update
* minor edits to nest backend and joint_test
* add spiketrain timescale test (decay of the autocorrelation function of a spiketrain)
* fix imports
* PEP8 style related updates
* PEP8 style updates
* fixing wasserstein_distance score
* add missing import and fix minor bug
* Fix indexing for differently sized samples
* modify return statement to provide the full nest object (where the kernel can be accessed)
* inherit from ProducesSpikeTrains
* remove merge markers
* fix iterables
* change file namespace and simplify __init__ files
* move nest import inside class
* remove prefixes in imports
* fixing import bugs
* adding grouped avg correlation test
* return nan if there is only one spiketrain in the list
* account for the possibility of nans
* include opencv to the environment
* fix bug in setup.py
* integrate grouped spiketrains into ProducesSpikeTrains
* handling empty input to wasserstein score
* Add std correlation test
* remove np.squeeze
* fix authors textfile
* include new parameter to control for number of spikes before calculating the timescale
* PEP8 issues
* Include LvR to variation tests
* PEP8 compliance
* actual name setting in the test instance
* update requirements
* comment out property leading to problems
* add power_specturm test
* improve power spectrum test
* typo in power spectrum test
* get ps test ready to roll!
* use the mean psd across groups
* add freqband_power test, make joint test more robust
* joint test example, small edit of avg_corr _test
* fix some imports and their style
* Incorporate changes from Jasper
* improve PEP8
* remove unused import
* include ability to deal with name as a parameter
* NeoHdf5IO is deprecated in the latest neo version. Updated to usign the NixIO
* introduce option to control the normalization of the observations and predictions to the Wasserstein distance
* fix standard deviation not being used!
* create test to calculate multiple metrics from the CC matrix at once, significantly reduces runtime
* ensure parameters exist
* ensure parameters exist
* remove merge marks
* many fixes
* remove test parameter setting downstream in favor of default_params upstream
* rename binsize to bin_size
* Create decorator for generate prediciton function and update all functions to avoid supefluous repetition
* minor docstring update
* update eigenanle score
* add context manager for easier parameter passing
* remove hard-coded default
* deploy keyword argument filtering context manager
* remove funky correlation test
* move parameter to default
* fix parames reference
* add temp release notes file
* remove parameter kwargs from all functions except the __init__ and rename the prediction decorator
* Improve parameter documentation consistency and fix a few self.params
* improve set_prediction
* adding parallelization utils
* inherit default_params in child test classes
* parallelize of spiketrains and lists of spiketrains
* small parallelization edit
* parallelize joint test
* generalize use_cache decorator, enable sharing of correlation matrix calculation across tests
* use cache for correlation matrix
* addon commit to caching for correlation matrix
* update and clean-up the use of backends
* add renaming also to __init__
* fix import bug
* remove outdated import
* add missing self.default_params
* merge tests and clean the code up
* update test imports
* enable caching and parallelization in the correlation test and joint test
* do not pass parameters to runnable model
* fix default params bug in models
* add missing model argument in correlation_matrix_test
* small bug fixes
* fix parameter handling in stochastic activity model
* fix name
* adding release notes
* fix proxy load in stochastic model; fix grouped sts check in joint test; fix import in isi var test
* better integrating model params and default params
* adapt wasserstein distance for non joint-test use
* making wasserstein distance symmetric by default
* fix typo
* fix bug for wasserstein 1D array input
* fix parameter handling in model classes
---------
Co-authored-by: Robin <robin.gutzen@live.de>
Co-authored-by: Aitor <a.morales-gregorio@fz-juelich.de>
Co-authored-by: Aitor MG <43403140+morales-gregorio@users.noreply.github.com>
Co-authored-by: morales-gregorio <aitormorales95@gmail.com>
