« first day (873 days earlier)      last day (3811 days later) » 

03:28
@LordStryker Nope, no math mode, even if it is very tempting. Everybody goofs now and again, sometimes I come across my own edits and it shoots to my mind: "Mother of Dragons, what was I doing here?!"
04:01
2
A: Is mathjax in titles a problem? And why?

Tim PostIt can be rather problematic, and we ask that you do it only when the extended markup is really needed for the meaning of your question in the title. That might be quite often, depending on the questions you ask, we just ask that you consider it. Titles are a major part of how search engines ind...

 
3 hours later…
06:49
0
Q: What happened to locked answer?

RutvikSutariaI have recently seen the answer which is locked. What happen to that answer, does it get delete out? Or get opened after some editing by moderators?

 
4 hours later…
11:07
@Martin thanks for the tip. I must have missed that memo somewhere. Its easy to put chemical formulas and such in 'text mode' but what about question titles that have actual equations in them?
11:35
@LordStryker I always try and go for a more descriptive title, that does not include a formula. In some very rare cases where it is not possible exceptions to the rule are also allowed. In most cases however, you can perfectly transform the title with a few twitches. In the end it should be about readability...
@Martin Sure thing boss. Sounds good.
@LordStryker Sometimes it's a real bummer and if I do not even have a clue what some people are asking. Usually the biggest problems make homework questions, where you just would exceed the title field until you get a meaningful question in there. But I figure it is very important, especially when you are later starting to look for duplicates. It's like 'I know there was a question, but I cannot think of the title...' and a very descriptive title comes in handy then...
Absolutely. Personal opinion is that HW questions are littering our site :(
Also, is CCSD a 'gold standard' where you're from? I might be crazy but I only ever heard of "CCSD(T)" being gold standard.
@LordStryker When did I write that??? I must have been lazy...
oh no not you, someone else
11:47
My 'gold value' standard is still DF-BP86/def2-SVP ;)
the '@Martin Standard' ;)
Well, well, well... it gives you reasonable results in the minimum time required. And after you have initial results you can go all brute force with the ab initio
Why not HF/STO-3G then? :P
minimal time! :)
@LordStryker Btw. Have you ever heard of ab initio DFT? I believe I read it somewhere, but could not really get my head around it.
Well, it depends. A lot of people argue that DFT is indeed ab initio while others seem to disagree. A semantical argument to say the least.
11:52
I would always prefer pm6 over HF/sto-3g, even faster. And you don't even need a basis set. I don't really like the HF approach, DFT at least tries for some coulomb correlation
I never use HF for anything other than test jobs.
Is this a new form of DFT that is under development @Martin?
@LordStryker Oh myyy... I think Density Functional Theory is complete and at infinity exact, but not the functionals we usually use.
@LordStryker That's what I was asking you... I have no clue. I believe DFT is the hunt for the golden grail :D
Ah ha, okay fair enough. But since all the DFT methods are parameterized (i.e. all implementations of DFT suck, er... are fitted to something), then DFT in general is pretty much non ab initio in my opinion.
I have also heard that DFT should nowadays be more likely referred to as Density Functional Techniques, as of the many errors and error cancellations
I will check around. My curiosity is peaked about this ab initio dft
I like that actually
11:56
That's what I was going for..
Yeah, if you caan get rid of the fitting that would be awesome...
but then again it might be as complicated as CCSD(T) and we are not gaining anything...
Maybe Donald Truhlar, our DFT lord and savior, will eventually save us.
I guess he will be parametrized to death :P
Exactly, even the cheap methods now need to be calibrated with wavefunction methods (if done properly) completely going against the spirit of DFT in the first place :(
All hail Truhlar.
2
Q: Dipole moment of cis-2-butene

yolo123I need someone to back me up on this before I go confront my teacher: I was doing some analysis of the dipole moment of cis-2-butene. Let's say that the alkyl groups are both on top. Would the dipole moment be a vector that has its tail on the top and its head on the bottom? How would one calcul...

I'm attacking this as a mini research project
whoops
coincidentally, I'm using a slew of truhlar's functionals for this. XD
I do not like the minnesota functionals, they afford almost the same time as MP2 and you can get b2plyp and stuff...
Eh, they can be pretty good for systems which have significant dispersion components to them.
MP2 sucks with that sometimes.
of course the Hessian index from DFT is seriously problematic.
That paper in summary, DFT does good with geoms and energetics with respect to MP2 and CCSD(T), respectively. Massive discrepancies in Hessians, especially for sulfur containing systems.
But I remember M06-2X doing pretty good. TPSS and VSXC pretty much fail across the board.
12:03
'MP2 is a graveyard for transition metal chemistry' I don't know who said it, be s/he is right...
I am home now, I might come back to that when I am at work...
Oh god I hate transition metals. I avoid them like the plague!
@LordStryker I understand how you can be intrigued to run these calculations, it is tempting :D
I know right?! I love these types of questions. Like when someone asked about intramolecular bonding in hydrogen peroxide (?) and I busted out some NBO.
I certainly learn a lot from it too.
@LordStryker I agree, they do propose a lot of troubles. Unfortunately I am working on copper catalysis... So DFT it is, even if it is right for maybe the wrong reasons
(I'm working with Pt so I'm forced to use B3LYP and some pseudopotentials)
goes against my very core but oh well
maybe I'm just 'selling out'
12:09
@LordStryker That's the thing, isn't it? You really learn how to apply these analysis methods again. Like using AIM and NBO and stuff
Absolutely. Plus I get to apply them in ways that i wouldn't have thought of before. I really really really like these kinds of questions.
I just hate B3LYP... it's just so sloppy, takes a lot of time and it is way more popular than it should be...
@Martin Thats what it gets for being 'the first' and 'widely available' functional
Whenever I am tempted to use B3LYP, I use PBE0, TPSSh and M0X first... then see if anyone wants to argue it is necessary to use that gaussian functional
@Martin I use it just because I know people will want to see it. But honestly it doesn't take that much time to do if I'm already busting out a whole slew of functionals.
12:12
On a conference (or talk), if anyone uses only B3LYP I am always tempted to ask: 'Have you verified your results to some other, more reliable functionals?'
@LordStryker Sure I understand, but it would never be the first choice, would it?
Any functional needs to be verified by either a shit ton of other functionals or MP2/CCSD(T)
ehhhhh, not a first choice no. Its usefulness only lies with its pervasive nature in the literature.
to me, "b3lyp" is just one extra string in a list of strings. All the effort is handled by the computer once I hit "go" on my script. :)
You seem to be lucky with a lot of pc power. I have only twenty cores and to big molecules. I can only afford bp86 and have to verify that against other functionals... too bad. I always have to be like, can I skip that?
12:36
@Martin I feel ya. Yeah, I'm really lucky indeed to have access to this kind of computer resources. You should move to the southeast (US)! We've got a massive concentration of computing power!
12:50
@LordStryker Hehe, I am actually quite happy here. Whenever I have not gotten any good results I can blame it on the computers...
 
6 hours later…
19:09
muwahahahaha 400 students, exam 1, half fail. 20 A's.
 
2 hours later…
21:27
@Martin Feel free to clean up any errors you spot. I'll be sure to do a clean read-through tomorrow.
0
A: Dipole moment of cis-2-butene

LordStrykerMini Research Project Time Preamble This response is in no way meant to be contrary to what Geoff has already posted. I happen to enjoy these types of questions and I like to tackle them as little research projects for fun. Therefore, my answer will be brutally detailed and in-depth. That sa...


« first day (873 days earlier)      last day (3811 days later) »