On 17/10/2016 00:41, Rick Walsh wrote:

Hi,

Sorry for being slow to give much of a response.  I started writing a
response but didn't quite finish it.

I think it's great to get the heat map documented in a meaningful manner,
so thank you both for working towards it.

On 11 October 2016 at 21:04, Willem Ferguson <willemfergu...@zoology.up.ac.
za> wrote:

> On 10/10/2016 17:34, Robert Helling wrote:
>
>
> On 10.10.2016, at 17:22, Willem Ferguson <willemfergu...@zoology.up.ac.za>
> wrote:
>
> give me a more precise indication of the unclear areas?
>
>
> It was simply the shire number of references to part A, part B then again
> A then C. Maybe the abstract letters don’t help. Could you use more
> descriptive names every now and then?
>
> In the end, it’s not so complicated: The heat map show how much gas is in
> the tissues compared to the ambient pressure, all this as a function of
> time. Blue colors mean less gas in the tissue and thus on-gasing, green to
> red mean excess gas in the tissue to various degrees and thus off-gassing.
> Fast to slow tissues from top to bottom. Lots of yellow and red mean lots
> of seed for possible DCS. This is the general idea, the more detailed
> version should be obvious from the figures (I would hope). I would think
> being less explicit in the text would help (because the reader needs to
> digest less information).
>
> What do you think?
>
> Best
> Robert
>
> --
>
> The previous gas pressure graph was pretty precise and I could read the
> exact compartment gas pressures from the graph and relate them to the
> gradient factor (accepting there was no calibrated Y axis). I prefer
> information that is as precise as possible, therefore this preference. In
> contrast, for the heat map the colours are more qualitative, but I still
> prefer as precise an interpretation of the colours as possible. Therefore
> just a general notion of blue to red is not sufficient. For instance, if I
> have a deco stage with orange in it, I want to understand whether the inert
> gas pressures were very close to the M-value or not. If the heat map cannot
> provide that information, it would not serve any purpose for me. Therefore
> my strong preference for a much more explicit interpretation of the heat
> map. Because of this, it took me some time to get to terms with the heat
> map. In addition, if I have never seen a heat map and I am reading the user
> manual for the first time, I would like to have the full explanation, not
> only a notional summary. The user manual is most useful for divers using
> these facilities of Subsurface for the first time. If I were that diver, I
> would have preferred the 'full treatment'.
> I hope my approach appears sane.
>

I think both approaches are entirely sane.  I think I'm agreeing with
Robert in that I believe the greatest strength of the heat map is that the
user can see very quickly, in a qualitative manner, when the greatest
tissue gradient of a dive occurs, and which theoretical tissues experience
the greatest gradient.  Heat maps were used effectively in the Simon
Mitchell presentation as a way of demonstrating the effects of deep stops
vs no deep stops.  The heat maps were used to compare two different dives,
which was a powerful comparison without needing to assign numerical values
to the colour scale.  I attempted to illustrate the concept in the graphic
I added to the user manual in commit fbcf4b9bd32a362078c81272e0e96f
87f1d0f980.

But I also think that someone interested in the heat map would want to know
what the colours actually represent, and for this reason we ought to
include the values too.  The direct comparison between the colour scale and
the instantaneous tissue graph is very helpful, as they are representing
the same thing.

My suggestion would be:
- Start the section with a paragraph explaining what the heat map
represents and how it works (basically Robert's text), with an example of a
heat map
- Provide the comparison with the instantaneous tissue graph, and provide a
numbered legend for the colours (basically Willem's text and graphics
currently in master).

The legend could be a single graphic, comprising an enlarged version of the
colour scale you've already made, annotated with:
- label oversaturated / offgassing zone, and undersaturated / ongassing
zone down left hand side of the scale
- tick marks and numbers for % M value at for example 120, 100, 80, 60, 40,
20, 0 down right hand side of upper part of scale, labelled as "% M value"
or "Gradient Factor"
- tick marks and numbers for % inert gas value at for example, 100, 75, 50,
25, 0 down right hand side of lower part of scale, labelled as "% inert gas
saturation"

If I get a chance, I will create a colour scale legend like the above, but
feel free to beat me to it (or suggest a better idea).

This could either be separate from, or combined into, the comparison with
the instantaneous tissue bar graph.

On a related point, I'm starting to think the ongassing zone should be the
% inert gas saturation for whatever gas is being breathed, rather than
air.  What do you think?  Currently, a dive with deco on 100% O2 will show
"ongassing" for the fastest and slowest tissues, when in actual fact they
are offgassing.  I think this is misleading.

Cheers,

Rick

I would support your proposal. The description for the top part of the
colour scale should be precise. Something like "Percentage of the distance
between ambient pressure (top of light green section in A) and the M-value
(bottom of red section in A)". Referring to the GF may be confusing here
since the heat map does not include information on the GF.
For the bottom part of the colour scale, perhaps "Percentage of the
distance between surface equilibrium inert gas pressure and ambient
pressure". Your last paragraph, above, is entirely logical.

Would you be prepared to put together such a colour gradient? As long as
the captions for the top and bottom parts of the pressure scale are
precise, this could be part of the first figure. I am, however, not sure
that there is enough screen width to put all of that in one single figure.
So maybe a separate colour gradient scale is more efficient.

I have ambivalence in interpreting the two heat maps that you put together,
comparing dives with and without a deep stop. For the fast tissues, the
purple section at the top of the heat map just before surfacing is deeper
for the dive without deep stop. One would have expected the fast tissues to
have offgassed more efficiently with a deep stop. Conversely, the slow
tissues are ongassing more slowly on the dive with the deep stop: at 30
minutes slow tissue areas in green in the top heat map (no deep stop) are
purple on the dive with a deep stop and the very slow tissues at the bottom
are lighter blue in the dive with deep stop. These observations are
counterintuitive. Am I missing something?

I have made some changes to the text to simplify in the way Robert proposed
and also to be more precise in other places. I will submit them as a patch
so that you can see these.

Kind regards,
willem

On Mon, Oct 17, 2016 at 12:41 AM, Rick Walsh <rickmwa...@gmail.com> wrote:

> Hi,
>
> Sorry for being slow to give much of a response.  I started writing a
> response but didn't quite finish it.
>
> I think it's great to get the heat map documented in a meaningful manner,
> so thank you both for working towards it.
>
> On 11 October 2016 at 21:04, Willem Ferguson <
> willemfergu...@zoology.up.ac.za> wrote:
>
>> On 10/10/2016 17:34, Robert Helling wrote:
>>
>>
>> On 10.10.2016, at 17:22, Willem Ferguson <
>> <willemfergu...@zoology.up.ac.za>willemfergu...@zoology.up.ac.za> wrote:
>>
>> give me a more precise indication of the unclear areas?
>>
>>
>> It was simply the shire number of references to part A, part B then again
>> A then C. Maybe the abstract letters don’t help. Could you use more
>> descriptive names every now and then?
>>
>> In the end, it’s not so complicated: The heat map show how much gas is in
>> the tissues compared to the ambient pressure, all this as a function of
>> time. Blue colors mean less gas in the tissue and thus on-gasing, green to
>> red mean excess gas in the tissue to various degrees and thus off-gassing.
>> Fast to slow tissues from top to bottom. Lots of yellow and red mean lots
>> of seed for possible DCS. This is the general idea, the more detailed
>> version should be obvious from the figures (I would hope). I would think
>> being less explicit in the text would help (because the reader needs to
>> digest less information).
>>
>> What do you think?
>>
>> Best
>> Robert
>>
>> --
>>
>> The previous gas pressure graph was pretty precise and I could read the
>> exact compartment gas pressures from the graph and relate them to the
>> gradient factor (accepting there was no calibrated Y axis). I prefer
>> information that is as precise as possible, therefore this preference. In
>> contrast, for the heat map the colours are more qualitative, but I still
>> prefer as precise an interpretation of the colours as possible. Therefore
>> just a general notion of blue to red is not sufficient. For instance, if I
>> have a deco stage with orange in it, I want to understand whether the inert
>> gas pressures were very close to the M-value or not. If the heat map cannot
>> provide that information, it would not serve any purpose for me. Therefore
>> my strong preference for a much more explicit interpretation of the heat
>> map. Because of this, it took me some time to get to terms with the heat
>> map. In addition, if I have never seen a heat map and I am reading the user
>> manual for the first time, I would like to have the full explanation, not
>> only a notional summary. The user manual is most useful for divers using
>> these facilities of Subsurface for the first time. If I were that diver, I
>> would have preferred the 'full treatment'.
>> I hope my approach appears sane.
>>
>
> I think both approaches are entirely sane.  I think I'm agreeing with
> Robert in that I believe the greatest strength of the heat map is that the
> user can see very quickly, in a qualitative manner, when the greatest
> tissue gradient of a dive occurs, and which theoretical tissues experience
> the greatest gradient.  Heat maps were used effectively in the Simon
> Mitchell presentation as a way of demonstrating the effects of deep stops
> vs no deep stops.  The heat maps were used to compare two different dives,
> which was a powerful comparison without needing to assign numerical values
> to the colour scale.  I attempted to illustrate the concept in the graphic
> I added to the user manual in commit fbcf4b9bd32a362078c81272e0e96f
> 87f1d0f980.
>
> But I also think that someone interested in the heat map would want to
> know what the colours actually represent, and for this reason we ought to
> include the values too.  The direct comparison between the colour scale and
> the instantaneous tissue graph is very helpful, as they are representing
> the same thing.
>
> My suggestion would be:
> - Start the section with a paragraph explaining what the heat map
> represents and how it works (basically Robert's text), with an example of a
> heat map
> - Provide the comparison with the instantaneous tissue graph, and provide
> a numbered legend for the colours (basically Willem's text and graphics
> currently in master).
>
> The legend could be a single graphic, comprising an enlarged version of
> the colour scale you've already made, annotated with:
> - label oversaturated / offgassing zone, and undersaturated / ongassing
> zone down left hand side of the scale
> - tick marks and numbers for % M value at for example 120, 100, 80, 60,
> 40, 20, 0 down right hand side of upper part of scale, labelled as "% M
> value" or "Gradient Factor"
> - tick marks and numbers for % inert gas value at for example, 100, 75,
> 50, 25, 0 down right hand side of lower part of scale, labelled as "% inert
> gas saturation"
>
> If I get a chance, I will create a colour scale legend like the above, but
> feel free to beat me to it (or suggest a better idea).
>
> This could either be separate from, or combined into, the comparison with
> the instantaneous tissue bar graph.
>
> On a related point, I'm starting to think the ongassing zone should be the
> % inert gas saturation for whatever gas is being breathed, rather than
> air.  What do you think?  Currently, a dive with deco on 100% O2 will show
> "ongassing" for the fastest and slowest tissues, when in actual fact they
> are offgassing.  I think this is misleading.
>
> Cheers,
>
> Rick
>
>
>
>
> _______________________________________________
> subsurface mailing list
> subsurface@subsurface-divelog.org
> http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
>
>
_______________________________________________
subsurface mailing list
subsurface@subsurface-divelog.org
http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface

Reply via email to