expand CellResponse.mean_rates() to calculate binned spikes rates over time for plot_spikes_hist()

[rythorpe]

Based on our discussion in #640, we should consolidate code in plot_spikes_hist and CellResponse.mean_rates() when calculating binned spike rates over time. In addition to adding the functionality to calculate mean spike rates over time, this will fix the rare viz.plot_spikes_hist() error that pops up when there are no spikes in CellResponse.spike_times since the bins calculation depends on a non-empty list.

[jasmainak]

could this be a beginner or intermediate issue?

[rythorpe]

Not sure honestly. Whoever works on it will just need to think carefully about how to efficiently use the CellResponse object so that we make the code for each of the above functions more compact and efficient.