Statistics: Posted by tagoma — Today, 12:04 pm

]]>

There shall be a new RINs category called "E-RINs", "E" standing for "Electric", that may be included into the D3 RINs pool.

How all this novelties will work out is unclear at this stage .

But what is pretty clear -- if you ask me -- is that this will be more public money (we are talking about billion and billion bucks, here) going right into Musk's pocket.

Have a good day.

tags

Statistics: Posted by tagoma — Today, 11:37 am

]]>

Cuch, very sorry for delay in my reply. Below is the Mathematica code to solve any 1st order or nth order ODE. I am writing the code in text format below and explain it again in a next code window.

..

In Below I explain the above code with some comments..

I think your ODE(2) can be very easily solved. ODE(1) seems singular at t=0 but can possibly be solved by changing starting value to some positive time and specifying initial conditions there.

Please feel free to use the mathematica ODE solver above and come back with your comments.

Sorry, I have been pretty busy with work on my trading algorithms and want to apply them on NASDAQ stocks so little time left for other things. Will come back soon to tackle the discussion.

In my paper, I have given a mathematica routine that can solve arbitrary systems of nth order ODEs (at least in theory). I have also given an example explaining the algorithm where I used the algorithm to solve a system of three 2nd order non-linear ODEs.

I have non-hands on experience of Mathematica from a previous article with Alan and Paul..

Code:

`Clear[t,y,Zi,Wi,Yi,Y,y0,i,i1,p,n,ZAns];n=1;Array[y,n+1,0];y[1]:= y[0]^2;Array[y0,n,0];y0[0]=1;p=8;Y:=y0[0];For[k=1,k<n,k++,Y=Y+y0[k] *t^k/k!];Array[Zi,n+p-1,0];Array[Wi,n+p-1,0];Array[Yi,n+p-1,0];Array[ti,n+p-1,0];Zi[n]=(y[n]/.t-> ti[n]);For [i=n,i<p+n-1,i++,(Zi[i+1]=0;For [k=0,k<n,k++,(Zi[i+1]=Zi[i+1]+D[Zi[i],y[k]]*(y[k+1]/.t-> ti[i+1]));]);];For[i=n,i<= p+n-1,i++,(Wi[i]=Zi[i];For[k=0,k<n,k++,(Wi[i]=(Wi[i]/.y[k]-> y0[k]))];)];For[i=n,i<= p+n-1,i++,(Yi[i]=Wi[i];For[i1=i,i1>=1,i1--,(Yi[i1-1]=Integrate[Yi[i1],{ti[i1],0,ti[i1-1]}])];Y=Y+(Yi[0]/.ti[0]-> t);)]ZAns=Collect[Y,t,Simplify]//PolynomialForm[#,TraditionalOrder -> False]&`

In Below I explain the above code with some comments.

Code:

`Clear[t,y,Zi,Wi,Yi,Y,y0,i,i1,p,n,ZAns]; This line is handy to clear variables especially when you want to solve for more than one ODE.n=1; n is order of the ODE. For 1st order ODE n=1; Array[y,n+1,0]; Define a workhorse variable.y[1]:= y[0]^2; Specify the ODE. y[1]= dy/dt. y[0]=y[t] and similarly for higher order ODE's y[2]=d2y/dt2Array[y0,n,0];y0[0]=1; Specify initial conditions. y0[0]=y(0) similalry for 2nd order ODE, you will also have to specify y0[1]=dy(0)/dt and so on.p=8; Number of powers in expansion of series.Y:=y0[0]; Allocate first initial condition.For[k=1,k<n,k++,Y=Y+y0[k] *t^k/k!]; Allocation of higher order initial conditions.Array[Zi,n+p-1,0];Array[Wi,n+p-1,0];Array[Yi,n+p-1,0];Array[ti,n+p-1,0];Zi[n]=(y[n]/.t-> ti[n]);For [i=n,i<p+n-1,i++,(Zi[i+1]=0;For [k=0,k<n,k++,(Zi[i+1]=Zi[i+1]+D[Zi[i],y[k]]*(y[k+1]/.t-> ti[i+1]));]);];For[i=n,i<= p+n-1,i++,(Wi[i]=Zi[i];For[k=0,k<n,k++,(Wi[i]=(Wi[i]/.y[k]-> y0[k]))];)];For[i=n,i<= p+n-1,i++,(Yi[i]=Wi[i];For[i1=i,i1>=1,i1--,(Yi[i1-1]=Integrate[Yi[i1],{ti[i1],0,ti[i1-1]}])];Y=Y+(Yi[0]/.ti[0]-> t);)]; This is integration loop which iteratively integrates starting from zero. replace zero with initial time when it is different from zero.ZAns=Collect[Y,t,Simplify]//PolynomialForm[#,TraditionalOrder -> False]&`

I think your ODE(2) can be very easily solved. ODE(1) seems singular at t=0 but can possibly be solved by changing starting value to some positive time and specifying initial conditions there.

Please feel free to use the mathematica ODE solver above and come back with your comments.

Sorry, I have been pretty busy with work on my trading algorithms and want to apply them on NASDAQ stocks so little time left for other things. Will come back soon to tackle the discussion.

In my paper, I have given a mathematica routine that can solve arbitrary systems of nth order ODEs (at least in theory). I have also given an example explaining the algorithm where I used the algorithm to solve a system of three 2nd order non-linear ODEs.

AFAIR the routine

Statistics: Posted by Cuchulainn — Today, 11:27 am

]]>

Take your time. Don't want be a party pooper, but the mathematical and numerical issues are far from simple.

Statistics: Posted by Cuchulainn — Today, 10:37 am

]]>

I have this:

Code:

`thedictionary = {'price money': '$1', 'name': 'Google', 'color': '', 'imgurl': 'http://www.google.com/images/nav_logo225.png', 'charateristics': 'No Description', 'store': 'google'}cur.execute("INSERT INTO product(store_id, url, price, charecteristics, color, dimensions) VALUES (%d, %s, %s, %d, %s, %s)", (1, 'http://www.google.com', '$20', thedictionary, 'red', '8.5x11'))`

I'm not sure where to go from here. I couldn't locate anything related on the internet, but I did come across this page, which stated that "you may use register adapter() to adapt any Python dictionary to JSON, either registering Json or any subclass or factory building a suitable adapter." Is that right?cur.execute("INSERT INTO product(store_id, url, price, charecteristics, color, dimensions) VALUES (%d, %s, %s, %d, %s, %s)", (1, 'http://www.google.com', '$20', thedictionary, 'red', '8.5x11')) psycopg2.ProgrammingError: can't adapt type 'dict'

I'm not sure how to perform this precise thing, as I'm new to psycopg2.

Statistics: Posted by Piyushbhatt — Today, 8:39 am

]]>

Friends, I have decided to write 3-5 formal papers about the material in this thread that would ideally be submitted to Wilmott magazine and I will soon be approaching Wilmott people for their requirements and guidelines.

The research subject of first paper would be finding the dynamics of variance/volatility SDE so that volatility surface is closely fitted across time given that asset SDE in two dimensional correlated SDE system is given either by lognormal or CEV dynamics.

All through this paper, we will remain mainly in a single dimensional SDE framework other than just a few small excursions into two dimensions.

For my first paper, I am cleaning up my work on one dimensional SDEs with addition of cumulants. And also on time integral of the mean-reverting and other SDEs. I want to solve for distribution of integral of volatility/variance SDE by backing it out from option prices using lognormal or CEV formula. And then I will find the dynamics of volatility SDE so that it perfectly fits the implied distribution of integral of variance. While backing out integral of volatility, I will also have to do accounting for correlation and contribution of correlated part of SV SDE but this should be simple given the insights we have learnt about variances and their addition within Z-series framework. All of this will be done in a one dimensional SDE framework. We have already done 70% of the work in this thread. I will be changing a few things like I do not want to go into cumulants at all (even though they should be used a lot for inference) and will be using new alternative methods. I will also be making some changes to calculation of time integrals of SDEs. Only major part that remains to be done is bootstrapping of integral of variance SDE and correlated part of the SDE from the cross-section of option prices. Since we want to remain in the one dimensional domain, I really hope that computational time for option chains out to 4-5 years will be in seconds and not in minutes.

And my time frame to complete the unfinished part is a few weeks. I will be posting the code on Wilmott as usual.

Statistics: Posted by Amin — Yesterday, 3:21 pm

]]>

Goal

We wish to apply efficient and robust PDE methods for Asian options as described in

https://onlinelibrary.wiley.com/doi/epd ... wilm.10366

https://onlinelibrary.wiley.com/doi/epd ... wilm.10620

to problems that models the term structure of plain vanilla option volatility. This problem has been addressed by Haug, Haug and Margrabe (2003) based on discrete versions of the Turnbull-Wakeman of Curran approximations. The perceived benefits of using PDE models an the best finite difference schemes (as in Duffy 2022) are:

. Accurate and efficient FD schemes using ADE (Alternating Direction Explicit) .. typically a fraction of a second to compute option price.

. The same PDE model can be use for a range of Asian-style options.

. Model makes no assumptions on the range of values of the underlying parameters as in HHM (2003). For example, the model is stable and accurate for all values of the volatility.

. It is applicable to a wider range of problems than is possible with (semi) analytical methods.

. We only need to rewrite that part of the C++ application that computes volatility.

Thus, we use the formulae (Curran versus Levy) in HHM (2003) for the volatility term structure and we inject the computed volatility (is it a number or could it be a step-wise continuous function?) into the PDE model instead of using analytic methods to compute the option price.

Is this a feasible approach and what are the advantages compared with opportunistic, context-sensitive solutions? At first glance, the two approaches should give the same (almost) results.

Statistics: Posted by Cuchulainn — Yesterday, 1:39 pm

]]>

]]>

I've just been targeted by tyreextinguishers.com! They don't like my car. (If they ever find out I have four cars, one of which does 8mpg...!)

They take the cap off, put a lentil in and screw the cap back down. The air slowly escapes, tyre goes flat.

Their website very helpfully explains how to remove the cap in the first place, lefty loosey, righty tighty. Clearly they are as thick as planks.

And, yes, they really use a lentil! A lentil! Sorry, kat, but a lentil! I think that says everything you need to know.

Anyway, brilliant idea...removing the caps will mean they have to crouch down while the tyre deflates. Which they probably won't do. But that's not ideal for the health of your valves. So how about a dummy valve? It looks like an ordinary valve, protects the real valve like the cap, but confuses the heck out of people of low IQ.

(The regular anti-theft dust caps often use standard hex/allen keys, too easy to remove. I did think maybe smearing some dog poo...readily available at the side of the road...on the cap would be a quick solution.)

Who's in?

lentil - Copy.gif

Lentil in bottom right of photo.

Not a GREEN lentil -- ?They take the cap off, put a lentil in and screw the cap back down. The air slowly escapes, tyre goes flat.

Their website very helpfully explains how to remove the cap in the first place, lefty loosey, righty tighty. Clearly they are as thick as planks.

And, yes, they really use a lentil! A lentil! Sorry, kat, but a lentil! I think that says everything you need to know.

Anyway, brilliant idea...removing the caps will mean they have to crouch down while the tyre deflates. Which they probably won't do. But that's not ideal for the health of your valves. So how about a dummy valve? It looks like an ordinary valve, protects the real valve like the cap, but confuses the heck out of people of low IQ.

(The regular anti-theft dust caps often use standard hex/allen keys, too easy to remove. I did think maybe smearing some dog poo...readily available at the side of the road...on the cap would be a quick solution.)

Who's in?

lentil - Copy.gif

Lentil in bottom right of photo.

Disappointing.

Statistics: Posted by Marsden — Yesterday, 10:53 am

]]>

Baguettes, a wine festival and Paris rooftops vie for a spot on UNESCO heritage list

Universal indeed! Just like the Champs Elysées, Saint Michel and old Beaujolais wine.

Universal indeed! Just like the Champs Elysées, Saint Michel and old Beaujolais wine.

French baguette may be included in UNESCO heritage list

Statistics: Posted by tagoma — November 29th, 2022, 11:13 pm

]]>

Statistics: Posted by Cuchulainn — November 29th, 2022, 9:47 pm

]]>

BTW I wonder how Phelim Boyle got the brainwave to price options using MC?

Statistics: Posted by Cuchulainn — November 29th, 2022, 6:22 pm

]]>

Statistics: Posted by tagoma — November 29th, 2022, 6:05 pm

]]>

Of course people pimp up their photos. Do you think I’m really as attractive as I look in mine?The number of attractive young women on LI holding a PhD, working in the HR space, who have never talked to me before but absolutely want to know what I think about their personal projects (yes all that at once) is just insane.

Statistics: Posted by Cuchulainn — November 29th, 2022, 5:51 pm

]]>

In Joburg, parked Bentleys get a dedicated guard.

Statistics: Posted by Cuchulainn — November 29th, 2022, 4:56 pm

]]>