Under review

Statistics > Paired Samples

Peter Ronning 9 years ago in General updated by Steve Kuterescz (Moderator (EN)) 9 years ago 2
Continuing to venture into new (to me) territory within MICROMINE, I took a look at the "Paired Samples" utility.  I am in fact working with such paired data, using files created by a colleague using a competing software package.

Reading the help, I found myself confused by the following:

"For instance:

"If there is a paired interval between hole AZ123 From 7 To 8 and hole BQ456 From 9 To 10, then there will not be an entry in the output file showing a pair between hole AZ123 From 7 To 8 and hole BQ456 From 9 To 10.

"One sample in a hole may however be paired with several samples in an adjacent hole such as hole AZ123 From 10 To 12 paired with Hole BQ456 From 15 To 16 and paired with Hole OF543 From 6 To 7.
"Adjacent intervals in a hole will not pair with each other but may pair with the same intervals in different holes. Such as Hole AZ123 From 6 To 7 paired with hole BQ456 From 8 To 9 and hole AZ123 From 7 To 8 paired with hole BQ456 From 8 To 9."

The first paragraph under "For Instance" seems to say something like "If there is X, then there will not be an entry in the output file showing X".  Seems kind of self-cancelling, like "If a pair is found, it won't be recorded."

I think it's just intended to explain that duplicate records won't get written to the output file, but that isn't really what it says.  Is that what is intended?
Under review
Hi Peter,

For now the most important thing is ensuring you get the result you expect, irrespective of the documentation. If so then I'll ensure the help is updated.

My understanding of the text is:

Paragraph 1: If a pair is made from hole A to hole B, the reverse pairing of the same samples (i.e. from hole B to hole A) won't be written to the file since that would duplicate the original pairing.

Paragraph 2 seems to be saying that a sample in one hole may be paired with any number of samples from a adjacent holes.

Paragraph 3 seems to be saying that adjacent downhole intervals in the same hole will never be paired with each other.

I think the use of actual hole IDs and sample intervals without a corresponding diagram is making the text more complex than it needs to be. I hope this clears things up.
Yes, I believe that is the intention. Thanks for your very useful feedback. The content certainly needs to be clarified.