SERVING THE QUANTITATIVE FINANCE COMMUNITY

Search found 1027 matches

  • 1
  • 2
  • 3
  • 4
  • 5
  • 69
by AVt
February 3rd, 2015, 8:04 pm
Forum: Brainteaser Forum
Topic: exp(5) = [$]e^5[$]
Replies: 545
Views: 124746

exp(5) = [$]e^5[$]

<t>Yes, we have different attitudes for that, here I would prefer to havea good initial instead of a generic approach or naming methods. And soavoiding probably expansive methods.Since generations already considered that problem (and the best I haveread about is '2 steps') here is the one I like wit...
by AVt
January 31st, 2015, 8:02 pm
Forum: Brainteaser Forum
Topic: exp(5) = [$]e^5[$]
Replies: 545
Views: 124746

exp(5) = [$]e^5[$]

<t>I do not mind ad hoc. There where fixed parameters, no?I agree to reject complex numerics for usual code. But use order=3 in E=0, 0 <= M < 1/3. Solving for the (first) real solution will do as guess, it should be X0 = rho/eps + 2*(eps-1)/rho where rho = ((3*M+(-12/eps^2*d)^(1/2))*eps^2)^(1/3), d=...
by AVt
January 31st, 2015, 4:07 pm
Forum: Brainteaser Forum
Topic: exp(5) = [$]e^5[$]
Replies: 545
Views: 124746

exp(5) = [$]e^5[$]

Either allow complex numerics or use the absolute value of X0.
by AVt
January 31st, 2015, 2:04 pm
Forum: Brainteaser Forum
Topic: exp(5) = [$]e^5[$]
Replies: 545
Views: 124746

exp(5) = [$]e^5[$]

The initial guess X0 = Pi/2 + (sqrt(1+eps*(2*(M+eps)-Pi)) -1)/eps should do for Newton (using approximation of order 2 in E = Pi/2). Not quite a task I want to do by paper and pencil.
by AVt
January 4th, 2015, 8:09 pm
Forum: Programming and Software Forum
Topic: A faster exp() finction
Replies: 10
Views: 4872

A faster exp() finction

Thx. So it is to get exp correctly up to the last (binary) digit, rounded correctly (convention whatever), IEEE 754 for C (AFAIK C++ does not officially guarantee IEEE 754)
by AVt
January 3rd, 2015, 9:18 pm
Forum: Programming and Software Forum
Topic: A faster exp() finction
Replies: 10
Views: 4872

A faster exp() finction

I do not quite understand "8* speedup". Is it for the multiple precisionin glibc or for the usual IEEE-754 double precision? And I miss somethingtowards correctness (see "The road ahead" at the end of the link). Hm ?
by AVt
December 29th, 2014, 7:47 pm
Forum: Numerical Methods Forum
Topic: iv for all and all for iv
Replies: 102
Views: 16392

iv for all and all for iv

<t>I thought the sketched would be enough ... [at extreme vol] 1) classical codesuffers from cancellation and 2) numerically an inverse is neither uniquenor discrete. Except you allow for arbitrary precision.1) For the example K=70 classical code would need 18 + 2 decimals to getcorrect results in 1...
by AVt
December 27th, 2014, 7:54 pm
Forum: Numerical Methods Forum
Topic: iv for all and all for iv
Replies: 102
Views: 16392

iv for all and all for iv

<t>Just some remarks (Spot = 100, Rates = 0, vol = 0.04% for data ?): It does not make much sense to "solve" for volatility if one does nothave a good pricer, which solver ever is used.One can use systems like Mathematica or Maple with high precision forextreme values (if done with care). But fixed ...
by AVt
March 7th, 2014, 9:07 pm
Forum: Technical Forum
Topic: Crypto problem
Replies: 62
Views: 8425

Crypto problem

<t>I would (theoretically ) distinguish goverment and publics. There are some waysand extending Tor to 'cover' the IP, using mixed cascades through providers whoassert they do not store (say AN.ON for http(s) or the german CCC for mails)and some who are small enough to be not forced to by (german) l...
by AVt
March 6th, 2014, 8:57 pm
Forum: Technical Forum
Topic: Crypto problem
Replies: 62
Views: 8425

Crypto problem

<t>Will try tomorrow, classical only (off "new" media)."Einen Staat, der mit der Erklärung, er wolle Straftaten verhindern, seine Bürger ständig überwacht, kann man als Polizeistaat bezeichnen. Den Polizei- oder Überwachungsstaat wollen wir nicht. Aber wir wollen, dass der Staat seine Sicherheitsauf...
by AVt
March 6th, 2014, 6:41 pm
Forum: Technical Forum
Topic: Crypto problem
Replies: 62
Views: 8425

Crypto problem

<t>I still do not understand what actually is desired (where I believe in the rule of thumb: one can make it "expensive" only). Is it about increasing privacy? Then Tor is a practical and easy way.But if you allow that an access to your data container can be observed then that may be a problem (even...
by AVt
January 1st, 2014, 6:43 pm
Forum: Student Forum
Topic: Variance Gamma - different results
Replies: 45
Views: 9903

Variance Gamma - different results

Thx! Switching the sign on input gives the desired value, 11.3700278104497 is whatthe above Excel+DLL returns (speed is roughly 5000 prices/second, 14 decimals), andK=90 --> 19.0993547242021, K=120 --> 1.92109238898072
by AVt
December 31st, 2013, 6:39 pm
Forum: Student Forum
Topic: Variance Gamma - different results
Replies: 45
Views: 9903

Variance Gamma - different results

Good!So what is the corrected version (in terms used in the paper, either the Thm or the appendix) ?
by AVt
December 30th, 2013, 8:06 pm
Forum: Student Forum
Topic: Variance Gamma - different results
Replies: 45
Views: 9903

Variance Gamma - different results

<r>Thx, that Excel+DLL gives 10.9651389219440Yes, it is which formula to take, mine was appendix (=picture ?) I have not searched deeper, just <URL url="http://www.wilmott.com/messageview.cfm?catid=4&threadid=14514but"><LINK_TEXT text="http://www.wilmott.com/messageview.cfm? ... d=14514but">http...
by AVt
December 30th, 2013, 5:10 pm
Forum: Student Forum
Topic: Variance Gamma - different results
Replies: 45
Views: 9903

Variance Gamma - different results

One needs the formula in the 'Appendix' of the paper.A DLL implementation with Excel interface and a short sketch is here, giving that 10.9651
  • 1
  • 2
  • 3
  • 4
  • 5
  • 69
GZIP: On