Flagging any triplicate that has a value > 0.50 +/- than any other would be alright, but currently it seems to be missing many cases like that and I haven't seen a pattern as to why. If I understand what that code is saying it does not appear to be comparing locs 0 and 2.
Ideally here are some examples as to what would be flagged more accurately.
1.0 / 1.5 / [2.0] : in this case even though 0 and 1 are in range and so is 1 and 2; 0 and 2 are out of range so 2 would be flagged as it is a larger value.
1.5 / [2.1] / 1.3 : 0 and 2 are within range but 1 is out of both ranges so 1 should be flagged.
2.1 / [1.5] / 2.3 : similar to the last one, this time 1 is the smallest value but is still flagged as the other two are in range of each other.
[3.4] / 2.6 / 1.5 : here none are within range so 0 should be flagged as it is the largest.
1.0 / 1.3 / 1.5 : everything is in range and nothing flagged.
From Cameron:
Flagging any triplicate that has a value > 0.50 +/- than any other would be alright, but currently it seems to be missing many cases like that and I haven't seen a pattern as to why. If I understand what that code is saying it does not appear to be comparing locs 0 and 2.
Ideally here are some examples as to what would be flagged more accurately.
1.0 / 1.5 / [2.0] : in this case even though 0 and 1 are in range and so is 1 and 2; 0 and 2 are out of range so 2 would be flagged as it is a larger value.
1.5 / [2.1] / 1.3 : 0 and 2 are within range but 1 is out of both ranges so 1 should be flagged.
2.1 / [1.5] / 2.3 : similar to the last one, this time 1 is the smallest value but is still flagged as the other two are in range of each other.
[3.4] / 2.6 / 1.5 : here none are within range so 0 should be flagged as it is the largest.
1.0 / 1.3 / 1.5 : everything is in range and nothing flagged.