## Alec Mihailovs

Dr. Aleksandrs Mihailovs

## 4470 Reputation

19 years, 269 days
Mihailovs, Inc.
Owner, President, and CEO
Tyngsboro, Massachusetts, United States

## Social Networks and Content at Maplesoft.com

I received my Ph.D. from the University of Pennsylvania in 1998 and I have been teaching since then at SUNY Oneonta for 1 year, at Shepherd University for 5 years, at Tennessee Tech for 2 years, at Lane College for 1 year, and this year I taught at the University of Massachusetts Lowell. My research interests include Representation Theory and Combinatorics.

## evalhf works with eval...

evalhf works with eval instead of subs. For example,

```A:=sin(x)+sin(x*sqrt(2)-1):
sol:=Optimization:-NLPSolve(A);

sol := [-1.56493598594498962, [x = -0.777194230546894]]

evalhf(eval(A,sol[2]));

-1.56493598594498962
```

This example is also interesting because it puts minimize(A); into a very long calculation - I couldn't wait until it finishes and interrupted it.

The answer given by NLPSolve is only a local minimum (closest to 0). A better value could be obtained specifying bounds for x,

```sol1:=Optimization:-NLPSolve(A,x=-1..5);

sol1 := [-1.85107897452703, [x = 4.26099489379575]]

evalhf(eval(A,sol1[2]));

-1.85107897452703218
```

However, if the interval is large and includes several local minimums, the answer given by NLPSolve may be wrong,

```sol2:=Optimization:-NLPSolve(A,x=-20..20);

sol2 := [-1.85107897452703, [x = 4.26099489606088]]
```

while

```sol3:=Optimization:-NLPSolve(A,x=16..20);

sol3 := [-1.99735092160169, [x = 17.3382042367515]]

evalhf(eval(A,sol3[2]));

-1.99735092160168758
```

as you can see, evalhf works pretty well here.

Alec

## evalhf works with eval...

evalhf works with eval instead of subs. For example,

```A:=sin(x)+sin(x*sqrt(2)-1):
sol:=Optimization:-NLPSolve(A);

sol := [-1.56493598594498962, [x = -0.777194230546894]]

evalhf(eval(A,sol[2]));

-1.56493598594498962
```

This example is also interesting because it puts minimize(A); into a very long calculation - I couldn't wait until it finishes and interrupted it.

The answer given by NLPSolve is only a local minimum (closest to 0). A better value could be obtained specifying bounds for x,

```sol1:=Optimization:-NLPSolve(A,x=-1..5);

sol1 := [-1.85107897452703, [x = 4.26099489379575]]

evalhf(eval(A,sol1[2]));

-1.85107897452703218
```

However, if the interval is large and includes several local minimums, the answer given by NLPSolve may be wrong,

```sol2:=Optimization:-NLPSolve(A,x=-20..20);

sol2 := [-1.85107897452703, [x = 4.26099489606088]]
```

while

```sol3:=Optimization:-NLPSolve(A,x=16..20);

sol3 := [-1.99735092160169, [x = 17.3382042367515]]

evalhf(eval(A,sol3[2]));

-1.99735092160168758
```

as you can see, evalhf works pretty well here.

Alec

## fun or not fun...

Well, you might be an exception that every rule must have, or there may be another rule actually - 10 comments unless you have another account.

Alec

## fun or not fun...

Well, you might be an exception that every rule must have, or there may be another rule actually - 10 comments unless you have another account.

Alec

## no code...

Code seems to disappear (as well as in many other old posts) - that's why I prefer posting the code, even if it is rather long, rather than providing a link to it - more chances that it won't disappear that way.

Pictures also disappear, even the recent ones, not even talking about the old 2D-math and MapleNet links.

Alejandro Jakubi posted recently a link to some web archive storing the old MaplePrimes. Perhaps, they (the code, pictures, and other links) could be restored from there if they were lost at Maplesoft?

Alec

## favorites...

I was just going over Joe Riel's posts adding some of them to my favorites, and noticed that. A strange thing is that the next in the list post (not empty) also had exactly 2131 views.

Alec

## Sage...

So you could just use it from Sage. It has the 1.1.6 linbox at the moment, but might update it in the not so far future. It is much easier than installing linbox separately - because Atlas is configured automatically during Sage installation, Givaro is already there etc. - much less trouble.

Alec

## Sage...

So you could just use it from Sage. It has the 1.1.6 linbox at the moment, but might update it in the not so far future. It is much easier than installing linbox separately - because Atlas is configured automatically during Sage installation, Givaro is already there etc. - much less trouble.

Alec

## Linux or Windows?...

Are you trying to do that in Linux, Mac, or in Windows?

In Linux or Mac, linbox is a part of Sage - so it can be used from Sage.

Alec

Thank you!

Alec

## Math Overflow, Gallager's paper, and exa...

So you asked that in the Math overflow and stackexchange, too? Perhaps, in some other places as well?

It would be a good idea to mention that here, I think, so that we could check the answers given there.

If you have some code written (in a small example), it would be also a good idea to post it here (in text form, so that it could be easier to copy and paste it in Maple.) Or just give a small, but complete example of the input and the output that you would like to have, at least including X1 and n.

If the Gallager's paper is available on the web, a link would be helpful. Otherwise, if you expect us to order it through the interlibrary access, the more detailed information is needed - the title and the issue number at least.

R. G. Gallager, “Low density parity check codes,” IRE Trans. Inform. Theory, vol. IT-8, no. 1, pp. 21–28, Jan. 1962.

Alec

## New bug...

It's a new bug on this site - if you are replying to somebody, and the reply has his/her (mine in this case) name at the top of it, then all what you see after submitting your reply is that name.

Try to post just as an answer, not as a reply to my comment.

Alec

## Declined...

It was considered and declined.

You don't have a right to appeal.

Alec

## For example?...

Can you provide an example? What branch cut would you like to have?

Alec

 1 2 3 4 5 6 7 Last Page 2 of 180
﻿