Using neighbouring values for plateau identification instead of the first value might bring about the occurence of
multiple peaks, that my not be present in the data, bringing about data inaccuracy. For example, data and epsilon as
input values for the peak function, if epsilon is large enough, multiple values can be identified to be equal,
therefore a plateau is realised.
If the plateau was searched using neighbours instead, the result might be multiple plateaus that are interpreted as
multiple peaks, even when this is not the case. The result would therefore be incorrect