mathematica.stackexchange.com/q/112125/121 looks to me like a duplicate of mathematica.stackexchange.com/q/83412/121 -- would a few others take a look at it?

Also mathematica.stackexchange.com/q/114085/121 appears to be a duplicate of mathematica.stackexchange.com/q/21861/121

Does anyone else find that some chemicals in ChemicalData have flat diagrams but not the nice 3D "MoleculePlot" option? e.g. ChemicalData["VanadiumIIIOxide", "MoleculePlot"] gives Missing[NotAvailable] but ChemicalData["VanadiumIIIOxide"] shows the flat molecule diagram.

Both v9 and v10, even after refreshing the curated data. Turns out the only thing with Vanadium that has a working MoleculePlot is "VanadiumIIIBromide"

@Verbeia A lot of the drugs in ChemicalData[] do not have the nice 3D structures. ChemicalData["Sildenafil", "MoleculePlot"] returns Missing["NotAvailable"], for instance.

@J.M. I wonder if they have any plans to expand the list or if there is a way to "create" the 3D plots from the 2D structure diagrams.

It might be possible to use EdgeRules and ElementTally with GraphPlot3D? Hmmm, maybe I have a question for the main site for the first time in ages.

There is a way to determine coordinates using VSEPR theory, but that takes more work. Most chemical software will do it for you; Mathematica is not one of them. :D

@Verbeia Well, with 3D, you'd need to find a database that has the appropriate files (*.mol, etc.) The coordinates have to come from somewhere.

I just tried GraphPlot3D[ChemicalData["VanadiumIVOxide", "EdgeRules"]] and it looks boring :-D

@Verbeia Method -> "SpringElectricalEmbedding" would be the poor substitute for VSEPR, since it at least tries to make the nodes repel each other.

@J.M. Isn't that the default?

@Verbeia Huh, so it is! Clearly I haven't used the thing in a while.

@J.M. thanks anyway - I've posted a question, as I don't have time to think about this now.

community.wolfram.com/groups/-/m/t/850968 - so let's change/add the extension of the main file type in WL and pretend nothing happened...

@J.M. With ref. to a/114065, I think WRI should extend FunctionInterpolation with Method -> "Chebyshev", since some of the machinery is in place. Since figuring out the above, I also got NDSolve to construct an accurate piecewise Chebyshev series for the unscaled Whittaker-root question.

NDSolve[..., Method -> "ExplicitRungeKutta", InterpolationOrder -> All] uses the built-in explicit RK 9(8) pair to construct order 7 series. I'm not sure why order 7. It doesn't seem to solve the DE in Chebyshev series à la chebfun, but probably converts the truncated Taylor series.

I'm musing over writing a chebfun style plugin for NDSolve. If time allows. Not sure how to get NDSolve to wrap the results up in a Chebyshev InterpolatingFunction. Might not be possible.

@Mr.Wizard The question is definitely closely related, however I don't think the question is a duplicate. Because all the matrices have dimension 2x2, the number of multiplications does not depend on the order I believe. Quite possibly it is a duplicate of this question (to which the other question by murphy was a follow up) instead, although that would contradict JasonBs answer in the (newer) duplicate.

@VitaliyKaurov As always with these historical things... I feel some of these are pretty speculative. Finnish history considers Kingdom of Finland a "project" instead of an actually existent form of government, and I have never personally heard of any historical evidence of Mongols in Finland, although they surely came close by...

@kirma Yep this is the reason I posted: first is to see errors in data - thanks for the hint. And second to discuss the limits of possible. See my own comment there on my post.

@VitaliyKaurov I feel many of such things are not necessarily true or false; rather they're a question of a perspective. Million such differing interpretations can exist, which makes life problematic for anybody seeking a bibliographic, "absolute" truth.

@DavidZ a familiar face