How can identify the incorrect data array from a graph?

I have a bunch of (30) data arrays (they are all 1 x 2218 arrays) and I have plotted them to show something like this:

In these plots it is clear to see the data arrays that do not follow the pattern (the purple, red and yellow ones that are lower than the other lines).

My initial idea was, after looking at the plot, to find all data arrays that had a value below 50 (for this specific case). But this would miss the yello and red pulse.

I have also thought of simply, dragging the mouse atop of these lines and identify specific xy-coordinates for these specific lines and have these found and eliminated, but this would be quite a slow technique and could be a bit problematic if the lines were too close to each other.

Is there any sort of filter I can apply, or another technique, that cleans all the incorrect data pulses/lines?

Answer 1

[Richard Burnside]

A very simplistic approach might to take the standard deviation of the all the data at each x value and and look for the outliers.