Benchmark Runs

[Div12345]

Functionality in eeg-nb to have like a benchmark run in the beginning. It would be useful, especially for beginners to have all the experiments run automatically and the decoding done via one of the known to be working algorithm and give metrics at the end, which would give an easy way for users to evaluate the functioning of the device and benchmark it with maybe previous ones stored somewhere, or even multiple benchmarks run by the same person with different devices or at different times.
If previously recorded decent quality data is present, it can be used to run a learning curve based on time length to identify the baseline time of the experiments to be run for the decoding to be good.

[Div12345]

This was the result for doing the learning curve for the P300 using MOABB's method. This should be easily extendable to other paradigms as well. @JohnGriffiths