Rounding up/down of maximum depths and other variables.

John Smith noseygit at hotmail.com
Wed Dec 9 15:07:03 PST 2015


Thanks dirk,  the numbers make sense when you know the thought process behind them.


________________________________________
From: Dirk Hohndel [dirk at hohndel.org]
Sent: 08 December 2015 15:28
To: John Smith
Cc: Subsurface Mailing List
Subject: Re: Rounding up/down of maximum depths and other variables.

On Dec 8, 2015, at 7:07 AM, John Smith <noseygit at hotmail.com<mailto:noseygit at hotmail.com>> wrote:

As I am comparing data between mobile and PC version, I am starting to understand more about how things are calculated, but have a few queries.

By the way, these are just questions rather than any thing else....

1. Why are depths rounded up/down to the nearest integer? With normal table use, the maximum depth is usually taken as the depth rounded to next highest integer.

What else would we do? I'm not quite sure what you are asking. We should report 11.1m as 12m? I don't think so.

2. Why are temperatures taken to the lowest value seen rather than the average over the whole dive - for example this:

It's a reflection of what most (?) people use the temperature value for: figure out what thermal protection to wear. It's also a left over from what older dive computers used to do: they stored just one temperature value for a dive and that was the lowest temperature measured (for many) or the temperature at the deepest spot (for some).

is recorded as a 27 degree dive, but you can see there were only a few moments when 27 degrees was recorded. Would an overall modal value be more representative?

3. And if you look at the dive list on the above grab, why do some depth values not get rounded whilst others do?

It's a matter of perceived precision (and certainly debatable). Showing 102.2m seems silly - actually, in general showing "10cm" resolution on depth is silly - is that your left arm, your right arm, or your toes? Yet not showing a decimal for single digits (6m, 7m) causes people to tell you that you are over-simplifying. So we compromise and show a decimal down to 20m (IIRC) and no decimal after that. I agree that this looks inconsistent - but I'm not sure what a more reasonable solution would look like.

/D



More information about the subsurface mailing list