Normalize cVEP datasets description columns

[PierreGtch]

• [x] To improve clarity in the doc, @thijor and I propose to change the docstrings and summary tables of cVEP (and eventually ERP) to these columns:

  =============  ======= ========= ============= ===== ============= ============== =============== ============== ===============
  Name           #Subj   #Sessions Sampling rate #Chan Trials length #Trial classes #Trials / class #Epoch classes #Epochs / class   
  =============  ======= ========= ============= ===== ============= ============== =============== ============== ===============

  With the following definitions:

• Trials length is the duration of one symbol selection

• #Trial classes is the number of different symbols

• #Trials / class number of trials per symbol in each session

• #Epoch classes for a visual cVEP paradigm, this would be the number of flash intensities

• #Epochs / class the number of flashes per intensity in each session

• [ ] And we should add those definitions to the dataset summary table

(see #557 and #562)

[PierreGtch]

I forgot the Presentation rate, frequency at which the flashes are presented

[PierreGtch]

Also forgot Codes