Fwd: Re: [help] Statistics Rework - first scratch.

[Willem Ferguson]

On 14/03/2016 03:03, Tomaz Canabrava wrote:
> All, Davide:
>
> This is by no means finished,
>
> Davide, I'm using your wireframe ideas on filtering, this screenshoot 
> uses fake data just to make it easier to present.
>
> What kind of graph you guys thinks it's the best?
>
Tomaz,

You are doing amazing stuff. Three suggestions on the statistics:

1) If only minimum, mean and average are plotted, then the statistically 
correct way to plot is by using three lines on the graph. A bar chart is 
inappropriate. With a bar chart, the vertical distance below the curve 
is meant to have content which it does not have in the present case. If 
one wanted to plot, for instance, "no. of dives with sac below 15", "no. 
of dives with sac between 15-15", "no. dives with sac above 25", then a 
bar chart would be appropriate. I include a mockup I made in R of what 
the appropriate chart should look like if one were to be statistically 
correct, using your min, mean and max.

2) The term "Average" is statistically undefined. It can imply any of 
several interpretations. E.g if I took 6 hours to drive the 300 miles 
from San Francisco to Los Angeles, the average speed would be 300/6 = 50 
mph.  However, if I measured my speed at 10 fixed points along the 
route, I am able to calculate (sum of all speeds)/10 =  mean speed. The 
appropriate term in our case is therefore "Mean".

3) In these graphs, the vertical axes need to be associated with a 
specific measurement unit (e.g. SAC (litres/min)). This brings up the 
issue of SI and imperial units.

Kind regards,
willem




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20160314/fe444941/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Screenshot from 2016-03-14 08:52:12.png
Type: image/png
Size: 31469 bytes
Desc: not available
URL: <http://lists.subsurface-divelog.org/pipermail/subsurface/attachments/20160314/fe444941/attachment-0001.png>