Make toy_example.py dependency on spikeextractors clearer

[h-mayorquin]

This is the first small step on to accomplish the goals described on #204. It is just changes that makes it clear what should be done to complete issue #204 .

[CodyCBakerPhD]

I was under the impression that the spikeextractors toy example utilized a lot of over-arching calls to things like generating waveforms and such. Are you saying we can just generate some dummy spike trains here and call it good?

[CodyCBakerPhD]

Though I wouldn't say it's the responsibility of our package to accurately generate biophysically realistic (or even statistically consistent) synthetic data. We just need example data structure to test and demonstrate usage that can be used in lieu of real data

[codecov]

Codecov Report

Merging #205 (e88468a) into master (5cff4ac) will decrease coverage by 0.09%.
The diff coverage is 0.00%.

@@            Coverage Diff             @@
##           master     #205      +/-   ##
==========================================
- Coverage   73.28%   73.19%   -0.10%     
==========================================
  Files          33       33              
  Lines        2276     2279       +3     
==========================================
  Hits         1668     1668              
- Misses        608      611       +3     

Flag	Coverage Δ
unittests	73.19% <0.00%> (-0.10%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/roiextractors/example_datasets/toy_example.py	6.66% <0.00%> (-0.28%)	⬇️

[h-mayorquin]

I was under the impression that the spikeextractors toy example utilized a lot of over-arching calls to things like generating waveforms and such. Are you saying we can just generate some dummy spike trains here and call it good?

Yep.

I agree that we should not care too much about realistic spikes. The toy example should be used for education / tutorial purposes anyway, right?

[CodyCBakerPhD]

The toy example should be used for education / tutorial purposes anyway, right?

Yeah, I believe the original motivation was the have 'realistic looking' output in the widgets applied to a demo file but... I don't know if they have to be THAT realistic looking... simple Poisson processes would be fine by me

[h-mayorquin]

Can we merge this? I just wanted to leave the terrain prepare for someone -maybe me later- to take this on and reference this from issue #204.