Animations/Blends of generalized Wagner; Hufnagel parameters

Discussion of troubles you experience and possible bugs.
Post Reply
Atarimaster
Posts: 446
Joined: Fri Nov 07, 2014 2:43 am

Animations/Blends of generalized Wagner; Hufnagel parameters

Post by Atarimaster »

Hello,

I just noticed something strange:
I tried to export animation frames.
Projection #1 was the generalized Wagner with default parameters, projection #2 was the generalized Wagner again using the parameters
a = 2.1418
b = 1.3128
m = 0.9492
m₂ = 0.843
n = 0.4444

No other attributes were changed, number of frames were set to 10.
In the exported TIFFs, frames 1 to 9 were identical, namely showing projection #1, and on the tenth frame it jumped to projection #2.

For testing purposes, I then selected the two maps again and selected "New Blended..." with P = 0.5. Again, Geocart just rendered projection #1, the information panel revealed that the parameters indeed were identical in projection #1 and the blended map.

On other parameterized projections (I tried Sinucyli, Strebe-Kavraiskiy V and Hufnagel) both blending and creating the animation frames works just fine.
So I guess there’s something wrong with blending two instances of the generalized Wagner?


And while testing this, I saw another oddity:
I used the Hufnagel interactive tool and selected Hufnagel X.
Then, I created a new map in Geocart using Hufnagel projection and copied the values given at the interactive tool (Ψmax = 30°, A = -0.6670, B =
0.6670, aspect ratio = 0.5) into Geocart’s parameter dialogue, but Geocart said: "ProjectionError: Projection parameters are not compatible with each other".

I tried a bit and saw that Geocart will render the projection when I set B = 0.666 or lower, but with B >= 0.667 I get that error message although the interactive tools allows values up to 1.

Kind regards,
Tobias
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

Re: Animations/Blends of generalized Wagner; Hufnagel parame

Post by daan »

Somehow I missed this posting, Tobias. Sorry about that. I will look into these matters.

— daan
Atarimaster
Posts: 446
Joined: Fri Nov 07, 2014 2:43 am

Re: Animations/Blends of generalized Wagner; Hufnagel parame

Post by Atarimaster »

No problem, and good luck! :)
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

Re: Animations/Blends of generalized Wagner; Hufnagel parame

Post by daan »

Atarimaster wrote:I created a new map in Geocart using Hufnagel projection and copied the values given at the interactive tool (Ψmax = 30°, A = -0.6670, B = 0.6670, aspect ratio = 0.5) into Geocart’s parameter dialogue, but Geocart said: "ProjectionError: Projection parameters are not compatible with each other".
I think Geocart is correct here. It’s impossible to see in the Hufnagel visualizer, but the mapped poles actually curl back below the latitude that has the greatest distance from the equator. Geocart never permits overlapping, and so it disqualifies this parameter combination.

A much more blatant example is [Ψmax = 90°, A = 0.2210, B = 0.8, aspect ratio = 2 (= 0.5 in Geocart)]. The visualizer permits this, but you can see that it results in a nonsensical topology. Geocart will not permit [B > ~0.546] in this circumstance. In the visualizer, it’s not clear at what B value the curling starts, but Geocart knows…

Cheers,
— daan
daan
Site Admin
Posts: 977
Joined: Sat Mar 28, 2009 11:17 pm

Re: Animations/Blends of generalized Wagner; Hufnagel parame

Post by daan »

Atarimaster wrote:So I guess there’s something wrong with blending two instances of the generalized Wagner?
Yes. It’s just a bug. It only affects a few projections. I’ve fixed it. Thanks for reporting it!

— daan
Atarimaster
Posts: 446
Joined: Fri Nov 07, 2014 2:43 am

Re: Animations/Blends of generalized Wagner; Hufnagel parame

Post by Atarimaster »

daan wrote: I think Geocart is correct here. It’s impossible to see in the Hufnagel visualizer, but the mapped poles actually curl back below the latitude that has the greatest distance from the equator. Geocart never permits overlapping, and so it disqualifies this parameter combination.
I see, this makes sense.
daan wrote: (Wagner)
Yes. It’s just a bug. It only affects a few projections. I’ve fixed it.
Great, thank you!
:)
Post Reply