187 replies to this topic

[Behnoud]

CPLua version 0.8 C. Is there any newer one?
LNA version 1.6, I think
Which instruction causes the error?
First I opened the file for example DBisect, and then pressed the Run on the above bar!
And about the erroneous instruction, I should say I didn't understand

[Orwell]

CPLua version 0.8 C. Is there any newer one?

Indeed, you should use CPLua 0.8 (final) instead.

When you see an error in a Lua program, you may press the key "<-" (the backspace arrow on the left of your keyboard). This will show you what caused the error.

[PAP]

CPLua version 0.8 C. Is there any newer one?
LNA version 1.6, I think

CPLua 0.8 C should run LNA version 1.6, but you should download the latetest CPLua version 0.8 anyway. To see what LNA version you have installed, run the program DemoAll (located at the directory LNAdrive), then select "About" from the main menu.
I'm 100% sure that there are no bugs in the example programs. If you cannot run any program in the LNAdrive or LNAexamp directory, you have probably deleted some LNA files by accident. If this is the case, you should reinstall LNA.

Btw, since you are a beginneer in CPLua, you should try the simpler examples at the LNAexamp directory first, and you really need to read the accompanying documentation. Try to undersand what each instruction does (Lua's documentation will also be a valuable help). Then ry to modify the example programs, to solve your own problems. Don't be too harry, and don't be dissapointed: you will surely make many mistakes at the beginning (everybody does that), but eventually you will realize that it's not as difficult as it seems at first glance.

Run the "driver" programs in the LNAdrive directory to see what LNA is able to do, but study them later, when you will know more things about CPLua. Driver programs are certainly not ideal for begineers.

[tsu]

Hi, i?m the RK annoying boy

I?ve looked de documentation, and with little time i don?t understand everything (except the "needs table").

PAP

I?m a 20 years student, doing chemical engineering, and the only things about programs taht i know are thisg VERY VERY VERY VERY basic from ocatve (imput, print, loops...). The friday i?ve got "Bifasic (biphasis?) reactors"?s exam, and for solve 70% of the problems, you need the RK method, cause you?ve got to solve two dif. equations at the same time.

If i?ve got more time, of course i?ll try to investigate how to do that operations, but I only have 3 days to study, and to learn the RK method (and practise it).

I surrender, i need the answer to that problem (example).

my differentials equations:

--> (dT/dX) = -65 + C1 / ((1-X) * exp (-C2/T))

--> (dt/dX) = 1/ (C3 * (1-X) * exp(-C2/T))

C1,C2,C3 are constants.

initials conditions:

--> X = 0 / T = 613 / t = 0

My teacher put the equations, solve them, and gave a table (X,t,T), with differents valors for T and t for each X (independient).

I suposse that a table with results or (better) equations T = f(X) & t=f(X) would be the correct output i need.

Please, explain me pass by pass (go to lua, go to that folder, open that, do that, change that, writte that...) how to get the solution of that equation sistem. Please please, i need it.

PD1: Why the Dsolve only solves easy linear equations? I put my equations and the answer (after tons of time) was the same equations.

PD2: I hate laplace. It doesn, t fit here, but i hate it.

[PAP]

Hi, i?m the RK annoying boy

Oh, no. You again (just kidding).

I surrender, i need the answer to that problem (example).

Hmm, you surrended too early. Normally, I should answer "try again", but, since you have only 3 days for your exams, I think I can help you.

my differentials equations:
--> (dT/dX) = -65 + C1 / ((1-X) * exp (-C2/T))
--> (dt/dX) = 1/ (C3 * (1-X) * exp(-C2/T))
C1,C2,C3 are constants.
initials conditions:
--> X = 0 / T = 613 / t = 0

Heavily nonlinear differential equations, but LNA should solve them easily. Give me C1, C2, and C3, and make sure that you have wrote the equations correctly.

My teacher put the equations, solve them, and gave a table (X,t,T), with differents valors for T and t for each X (independient).

Your teacher is not a magician. He uses numerical methods to solve them. Don't worry, you will do the same soon .

I suposse that a table with results or (better) equations T = f(X) & t=f(X) would be the correct output i need.

No, definitely no. Analytic solutions like T=f(X) and t=f(X) are impossible in Numerical Analysis. This is true for every computer system. However, you will get a table of (accurate) values.

Please, explain me pass by pass (go to lua, go to that folder, open that, do that, change that, writte that...) how to get the solution of that equation sistem. Please please, i need it.

Ok, ok. I will help you. Just send me the constants C1, C2, C3, and I will post the CPLua program for solving your problem. Don't tell me you expected such help, do you?

PD1: Why the Dsolve only solves easy linear equations? I put my equations and the answer (after tons of time) was the same equations.

dsolve is very poor. For systems of nonlinear equations, use my LNA function called Broyden. It's very powerful. See the example program XBroyden.

PD2: I hate laplace. It doesn, t fit here, but i hate it.

tsk, tsk, tsk. There is nothing wrong in Laplace transforms. In fact, it's a valuable mathematical tool. I'm afraid that you hate Mathemetics in general. If this is true, shame on you !

[tsu]

I twouldn?t be better a way to solve the problem (the method of imput) without de Cs valor? The problems have different Cs & expressions...


C1 = 1000 * 7,428 * 10^ (-19)
C2 = 22411
C3 = 1,937*10^(15)

in the solution-table, for example, for X=0,7:

T = 591 aprox
t = 13,8 aprox

PS (learning english): I hate laplace cause the june 6 i had the "instrumentation & control" exam. To solve the problems, with PID controllers, you needed to do laplace transforms (into s-domain), solve and do the inverse laplace tranf. . The exam goes so so, and that?s because i don?t like that.

[PAP]

I twouldn?t be better a way to solve the problem (the method of imput) without de Cs valor? The problems have different Cs & expressions...

Naive question. Yes, it would be better, but cannot be done. I'm not speaking about LNA only: every Numerical Analysis package cannot give a numerical solution for non-numerical constants C1, C2, C3. However, you can get a solution for each particular triplet of constants.

C1 = 1000 * 7,428 * 10^ (-19)
C2 = 22411
C3 = 1,937*10^(15)

in the solution-table, for example, for X=0,7:

T = 591 aprox
t = 13,8 aprox

Well, your values are not too accurate, especially for t(0.7). To be precise, the correct values are T(0.7)=589.760481, t(0.7)=15.471522, both accurate to 6 decimal digits . My RK4Rich function needs about 3 seconds to solve this problem . I can post a CPLua program that uses LNA to solve the problem, but first, let's talk about the price . I'll give you the program, but you have to say a nice proverb of your country first. I like to learn foreign proverbs. No proverb, no further help .

[tsu]

Naive question. Yes, it would be better, but cannot be done. I'm not speaking about LNA only: every Numerical Analysis package cannot give a numerical solution for non-numerical constants C1, C2, C3. However, you can get a solution for each particular triplet of constants.

Well, your values are not too accurate, especially for t(0.7). To be precise, the correct values are T(0.7)=589.760481, t(0.7)=15.471522, both accurate to 6 decimal digits . My RK4Rich function needs about 3 seconds to solve this problem . I can post a CPLua program that uses LNA to solve the problem, but first, let's talk about the price . I'll give you the program, but you have to say a nice proverb of your country first. I like to learn foreign proverbs. No proverb, no further help .

For example, there?s another excercise where the numerator of first equation is multipicated by (C4 - T). I guess it can be solved with de method you?re going to pass me, doesn, t it?

FAST EDIT: With the first part, i don?t refer to introduce Cs in the expression. I refer, por example:

IMPUT --> Equation 1 (writte any equation) Equation 2 (the same), contition, ....



Spanish proverb.

"M?s vale pajaro en mano que ciento volando".

The translation would be something like "better a catched bird, that a hundred flying"

Meaning: Is better to have something, despite being little (quantiti, cost,...), than try to have more and don?t get anything.

[tsu]

Going bed, it?s late here and tomorrow i must wake up early to study. Tomorrow i?ll see the answers.

Remember the "writte any equation" & "any conditions", pls.

Thanks for all, c u tomorrow

[PAP]

For example, there?s another excercise where the numerator of first equation is multipicated by (C4 - T). I guess it can be solved with de method you?re going to pass me, doesn, t it?

Don't worry, the program I have posted can be easily modified to solve similar problems.

Here you will find an image file containing LNA 1.60 and the program you need. In the "main" folder, just run the program "tsu". It solves the problem you posted, plots the results, then asks for a X-value and returns T(X) and t(X), iteratively. Enter a negative value for X to terminate the program. You will get something like this:

Now, if you want to modify the constants, change their values at the beginning of the program. If you want to modify the differential equations, you have to modify the function RHS in the program. It's very simple, even if you don't know programming: there is a couple of brackets {}, where the 2 differential equations are defined. Study how your current differential equations are written: x stands for X, y1 is the first function, i.e., T(X), and y2 is the second function, i.e., t(X). For example, if you want to define another constant, add C4=... and change the numerator of the first equation to (C4-y1), which means (C4-T).
Initial values are given in brackets where RK4Rich is called. Just change the values if you need to. In the call of RK4Rich you can also set the start and end points, and the accuracy of the result (in the program, the accuracy is set to 5E-4, which means 5*10^(-4), i.e. the result should have at least 3 accurate decimal digits). For example, to get a solution from x=0 to x=0.8 with an accuracy equal to 10^(-6), and with initial conditions T(0)=620, t(0)=0.2, change the call of RK4Rich to

y,xp,yp=RK4Rich(RHS, 0, {620,0.2}, 0.8, true, 1E-6)

I think that you will easily realize how it works. Let me know if you have further questions.

Spanish proverb.
"M?s vale pajaro en mano que ciento volando".
The translation would be something like "better a catched bird, that a hundred flying"

I understand the meaning (there is a proverb with the same meaning in my native language). Literally speaking, however, I prefer all birds flying, and I hate hunting and hunters.

[tsu]

I couldn?t sleep, forums are addictive (specially if you need some important content).....


AWESOME, AWESOME & AWESOME.

I don?t have words to describe your work. Simply awesome.

math.exp(thanks)

i?ve tried to run the program with different conditions. Tomorrow i?ll try to modifie ecuations. It seems easy, and your explanation about modifying is perfect.

I?ve seen in some problems 10^(....) , T^2 , log(T). I guess the two first ones are write with that symbol (^), and may be the third one written like "math.log(T)" ? Tomorrow trying ^^ Any guide to find how to write mathematical expressiones?

Again, thaks for all. Awesome.

[Orwell]

I?ve seen in some problems 10^(....) , T^2 , log(T). I guess the two first ones are write with that symbol (^), and may be the third one written like "math.log(T)" ?

That's right; the mathematical expressions are written in the "natural" way, you just have to add "math." before the name of math functions (e.g. math.sin, math.max, math.abs etc)

[PAP]

AWESOME, AWESOME & AWESOME.
I don?t have words to describe your work. Simply awesome.
math.exp(thanks)
i?ve tried to run the program with different conditions. Tomorrow i?ll try to modifie ecuations. It seems easy, and your explanation about modifying is perfect.

Wow, thank you. It's nice to know that I helped .
I have a feeling that CPLua's community has a new member now, and the small number of LNA users has been increased by one .

I?ve seen in some problems 10^(....) , T^2 , log(T). I guess the two first ones are write with that symbol (^), and may be the third one written like "math.log(T)" ?

Indeed, math.log(x) is the logarithmic function in Lua (natural logarithm), and math.log10(x) is... well, I think it's obvious.

Tomorrow trying ^^ Any guide to find how to write mathematical expressiones?

For a list of mathematical functions, look at section 5.6 of Lua 5.1 reference manual (this manual comes with the CPLua package in html form). You may also need to download the PDF version of Programming in Lua (here is Orwell's topic for CPLua downloads).

[PAP]

Coming soon: LNA 1.70

The animated image below shows some things (among others) you will be able to do with LNA 1.70 (it may take a few seconds to see the animation in normal speed, so be patient):



This screenshot shows two separate graphic windows, both animated. Top window: sin(a*x); bottom window: a*cos(x). Both functions are plotted for various values of the parameter a, within the interval [-1,1]. Note that this is not a trick; the image shows what you will see in your ClassPad, using LNA's improved LNAplot library . The Lua program that creates this double animation is... 20 lines long.

[Orwell]

Impressive

... But I doubt that those graphs really represent sin(x+a) and cos(a*x)

[PAP]

You are right, there was a small typo in my post: the first window shows sin(a*x), and the second shows a*cos(x). I have corrected my previous post. Thanks!

[PAP]

Try this link. If you cannot download LNA from it, there is probably an error in your Internet connection.

[PAP]

LNA 1.70 is almost ready; it includes a much more powerful LNAplot library, some new numerical methods, and named optional arguments (instead of positional). If someone is writing something serious based on LNA, I suggest to wait a few days, since new version permits to use LNA a lot easier than before. I'm currently writing the documentation, and I hope that I will be able to upload the new version within next week.

[omegavirus]

Nice!!! I'll wait for it because now I need to use the plot library, LNA is powefull and whit the documentation it's easy to anyone to use it, nice work PAP!!!

[Tchernoben]

I dunno yet what is LNA and I don't have enought time to read all the topic, then I think I'll down it and try

Could you mail/pm me to explain me fast what is it ?

[PAP]

I dunno yet what is LNA and I don't have enought time to read all the topic, then I think I'll down it and try
Could you mail/pm me to explain me fast what is it ?

LNA is a rather large software, consisting of many CPLua scripts; most of the LNA scripts are really compicated. Don't expect to learn how to use it in seconds; on the other hand, don't expect that it is really very difficult. In any case, you can't learn LNA by reading a short mail.
LNA comes with extensive documentation. The best way to learn LNA is to download and read the documentation; I think this will answer all your questions. . Furthermore, this topic has useful information about LNA; some other topics in UCF include examples of LNA in action.

[PAP]

Nice!!! I'll wait for it because now I need to use the plot library, LNA is powefull and whit the documentation it's easy to anyone to use it, nice work PAP!!!

Due to the new CPLua version released these days, I have changed several minor things in LNA, but now I have to take all screenshots for the documentation again. I was planning to release LNA 1.70 this weekend, but I'm afraid that it's not possible; it will take some days, maybe next weekend. Be patient, you won't be dissapointed .

[PAP]

Download LNA 1.70 ?

Thanks for your interest about LNA 1.70. My "deadline" for releasing version 1.70 was last weekend, but I didn't make it in time, as always.
Be patient, I will upload the new version in a few days...

[Casio Calculator Factory]

LNA 1.7, where?

[PAP]

I know, I know, you are right, the new version hasn't been released yet... I am way out of the initial plan. You won't believe it, but LNA 1.70 is ready about two months ago; however, the documentation hasn't been updated yet. Be patient, I hope I will find the time to finish the documentation soon.

[vanhoa]

Then dont upload your project to geocities, use esnips for example... Thanks!

[unique33]

I wanna a Numerical Methods package for Visual Fortran . I wanna to know that is the IMSL package the best one ?

and which platform provided by the www.vni.com I should use for windows ?

and any more recommandation ?

[PAP]

I wanna a Numerical Methods package for Visual Fortran . I wanna to know that is the IMSL package the best one ?
and which platform provided by the www.vni.com I should use for windows ?
and any more recommandation ?

First of all, welcome back, unique33.

Well, I think I can help you. There are several commercial packages, such as NAG and IMSL. In my opinion, there is no "best" package, and you should not pay for them. There is a huge library of Numerical analysis packages in the Netlib repository; they are all free, and they are very-very good, and well tested. Furthermore, look at GAMS, where you will find both free and commercial software categorized, so you will easily find the algorithm you want. Most of the packages in these sites are written in Fortran 77, but, if you want them in Fortran 90 (which is much better than 77), look at this web page, where you will find many packages translated in Fortran 90. I use packages from these sites for several years, and I had never had problems. Most of them are not simply good, they are excellent. Note that many commercial packages are based in free software you can find in the links I have suggested. I don't think that you will need to search elsewhere .

If you are new in Numerical Analysis, you will have one problem: there are many free packages for the same purposes, which one is preferable? Well, you will probably be satisfied with any of them, but here are some basic suggestions from a "veteran":
Matrix Algebra: The de facto standard for this huge topic is LAPACK. You will find routines for solving systems of equations (linear or not), computing the inverse or the eigenvalues of a matrix, and many many more useful things. Most commercial packages have simply copied this excellent package. It's written in Fortran 77, but you will find a Fortran 95 interface here.
Integration: Use quadpack, a very powerful package. It is able to compute virtually any definite integral accurately.
Ordinary differential equations: Try odepack. I don't think you will ever encounter a single initial- or boundary-value problem which cannot be solved by this package.
There are also many other free packages for solving more specific problems, such as function minimization, nonlinear fitting, etc. If you need any of them, let me know .

Btw, why Visual Fortran? it's not the best choice, definitely. You can find excellent free Fortran compilers and integrated developement environments (IDEs). For a free compiler in both Windows and <{GNULINUX}>, try gfortran or g95. For a free IDE, try Eclipse. If you are using <{GNULINUX}>, there are even more powerful solutions: try the Intel Fortran compiler, which is free for non commercial use, or SunStudio, which includes free C++ and Fortran 95 compilers, together with a very good IDE.

[unique33]

First of all, welcome back, unique33.

Thanks.

Well, I think I can help you. There are several commercial packages, such as NAG and IMSL. In my opinion, there is no "best" package, and you should not pay for them. There is a huge library of Numerical analysis packages in the Netlib repository; they are all free, and they are very-very good, and well tested. Furthermore, look at GAMS, where you will find both free and commercial software categorized, so you will easily find the algorithm you want. Most of the packages in these sites are written in Fortran 77, but, if you want them in Fortran 90 (which is much better than 77), look at this web page, where you will find many packages translated in Fortran 90. I use packages from these sites for several years, and I had never had problems. Most of them are not simply good, they are excellent. Note that many commercial packages are based in free software you can find in the links I have suggested. I don't think that you will need to search elsewhere .

If you are new in Numerical Analysis, you will have one problem: there are many free packages for the same purposes, which one is preferable? Well, you will probably be satisfied with any of them, but here are some basic suggestions from a "veteran":
Matrix Algebra: The de facto standard for this huge topic is LAPACK. You will find routines for solving systems of equations (linear or not), computing the inverse or the eigenvalues of a matrix, and many many more useful things. Most commercial packages have simply copied this excellent package. It's written in Fortran 77, but you will find a Fortran 95 interface here.
Integration: Use quadpack, a very powerful package. It is able to compute virtually any definite integral accurately.
Ordinary differential equations: Try odepack. I don't think you will ever encounter a single initial- or boundary-value problem which cannot be solved by this package.
There are also many other free packages for solving more specific problems, such as function minimization, nonlinear fitting, etc. If you need any of them, let me know .

Thanks for your recommendations .They are really helpful like always. I want the numerical packages for an engineering project ( my final project ) next term . The Fortran 90 Libraries are the best for me .

If I had this Numerical Packages this term ,I could sell my first programming request for more , but I wrote it using mathematica and I got less . just because it was not Executable and Just because the Mathematica is an Interpreter.

Btw, why Visual Fortran? it's not the best choice, definitely. You can find excellent free Fortran compilers and integrated developement environments (IDEs). For a free compiler in both Windows and <{GNULINUX}>, try gfortran or g95. For a free IDE, try Eclipse. If you are using <{GNULINUX}>, there are even more powerful solutions: try the Intel Fortran compiler, which is free for non commercial use, or SunStudio, which includes free C++ and Fortran 95 compilers, together with a very good IDE.

I have provided Visual Fortran 90 last month. I choose the "Visual" because my Project should have user Interface . If The user Interface was not nessecery I could use even the Cplua Emulator and The LNA Package for my project .

By Interface I mean real resizable windows , menus ,...

If we had Cplua not just Emulating the Classpad and if it was a real Lua Interpreter It was my first choice for my project . Because I,m know familiar with Lua language and because I ,m useing LNA Package for all of my projects .

I used The LNA Package for :
*Internal Combustion Engine Project :
" NASA Program " ( Romberg ,Brent , Broyden Methods used whitin this Project)

*Modern Control Engineering Project :
"Ploting the Respoce , Error to different Signals using Analytic and Numeric Approuch "

*Heating and Cooling the Building Project

but Unfortunatly we do not have Real Lua Interpreter .
If there is such a thing let me know.
thanks .

[PAP]

If I had this Numerical Packages this term ,I could sell my first programming request for more , but I wrote it using mathematica and I got less . just because it was not Executable and Just because the Mathematica is an Interpreter.

Like you said, Mathematica is an interpreter, therefore it is much slower than a Fortran executable (usually 2 or 3 orders of magnitude slower, sometimes even more).

I choose the "Visual" because my Project should have user Interface . If the user Interface was not nessecery I could use even the Cplua Emulator and The LNA Package for my project.

I usually write the mathematical program in Fortran 95. Then, if a graphical user interface (GUI) is needed, I write it in C++, using the QT or the GTK+ toolkits. You can easily mix Fortran and C++ codes in your program. However, implementing the GUI is a headache; it needs several hours of boring programming just for the show, not for the real thing.
If you only need to plot results, you can do it esily without C++ toolkits: Try the DISLIN Fortran plotting library (which is free for non-commercial use). It's an excellent library, able to produce impressive plots with little effort (all you have to do is to call plotting routines from your Fortran program). DISLIN also includes routines for creating a simple GUI.

If we had Cplua not just Emulating the Classpad and if it was a real Lua Interpreter It was my first choice for my project . Because I,m know familiar with Lua language and because I ,m useing LNA Package for all of my projects .

Keep in mind that the upcoming LNA 1.70 introduces named optional arguments, much like CPLua's packages. LNA 1.60's positional scheme for optional arguments is no longer supported. This means that you will need to do small modifications to all your programs based on LNA 1.60. This change is, of course, annoying for LNA users, but it was really necessary, since most LNA functions have lots of optional arguments.

but Unfortunatly we do not have Real Lua Interpreter .
If there is such a thing let me know.

The "real" Lua interpreter can be downloaded from this site. However, CPLua is different (and better, IMHO), so your CPLua programs won't work out-of-the-box. Especially LNA is heavily based on CPLua features not found in a normal Lua interpreter. I'm not planning to convert LNA for the real Lua interpreter, because this will be useful for writing Lua programs on the PC, and I will never do that; in my PC, I have several Fortran compilers installed, there is no reason to write a mathematical program in Lua.

[The_AFX_Master]

First, Happy new year to all forumers, and nice to see you again here unique33.

someone knows about CPLua 1.0? it will be released soon?. there a few monts since Orwell do not post here . btw, he deserve that vacations, because he did a lot of heavy work in 2006. But 0.9D is still with some problems, specially regarding speed an memory issues, i recently did a few tests of high memory usage on the CPLua 0.9d and OS 3, and lots of exception errors ocurred (adressing errors, said to me the CP), instead the traditional not enough memory message.. I still think that v0.8 is the most stable.

PD: Orwell can put a mode on the emulator that uses more RAM, 5-10MB (off course, is the PC). to heavy tasks?.. i mean, a CP3.0 Mode with the 384Kb RAM, a CP2.2 with 1080Kb, and PC without memory bounds. it's possible?. it would benefit the using of LNA for heavy tasks, as unique 33 said, CPLua is useful for pocket projects

[Orwell]

Don't worry I'm still here
I'm just really busy these days, and I don't have time for CPLua at the moment I hope I will be able to release CPLua 0.9 before the end of January (and CPLua 1.0 isn't far )

There are no memory bounds on the emulator... You can already use as much memory as you want
This is another reason why it is really difficult to find the potential memory bugs in CPLua, because such problems never happen on the computer

Oh, and happy new year to everyone

[PAP]

Don't worry I'm still here
I'm just really busy these days, and I don't have time for CPLua at the moment

Me too, I'm really horribly busy

I hope I will be able to release CPLua 0.9 before the end of January (and CPLua 1.0 isn't far )

I also hope that I will be able to release LNA 1.70 before the end of January.

I used The LNA Package for :
*Internal Combustion Engine Project :
" NASA Program " ( Romberg ,Brent , Broyden Methods used whitin this Project)
*Modern Control Engineering Project :
"Ploting the Respoce , Error to different Signals using Analytic and Numeric Approuch "
*Heating and Cooling the Building Project

it would benefit the using of LNA for heavy tasks

AFX_Master and unique33 (the two major funs of LNA) stike again. Thank you, lads. It's really nice to know that my project is useful to you .

Oh, and happy new year to everyone

Bonne annee! (the accent in "annee" is missing; I don't know how to type it in UTF-8)

[girdeux]

AFX_Master and unique33 (the two major funs of LNA) stike again. Thank you, lads. It's really nice to know that my project is useful to you .

Your project is very usful, I study chemical engineering, and I use CSpline, RK4Rich, Romberg, ... and other, but mainly this. And all of this is possible with CPLua, I have improbed my programs since I programing with CPlua. Thanks to the people that have made possible all of this.

Ahh happy new year too... feliz a?o nuevo... feli? any nou.

[unique33]

One of my friends gave me "Visual fortran compiler pro 9.1 with IMSL" as a gift . but It require "visual studio 2002 or 2003 ".

I wanna to know that if I provide the visual studio and if I install the fortran compiler , can I run and compile the Fortran 90 files through the visual studio ?

Also I need a good book for Fortran Programming .Any recommendation ?

Some books available on the print.google.com with limited preview :
. "Programmer's Guide to FORTRAN 90" By Walter S. Brainerd, Charles H. Goldberg, Jeanne C. Adams
. "Fortran 90/95 for Scientists and Engineers" By Stephen J. Chapman
. "Classical Fortran: Programming for Engineering and Scientific Applications" By Michael Kupferschmid
,....

I need the book for Engineering and Scientific Applications .

[unique33]

I successfully installed the compiler and the hello world project containing the library worked well.


How about the "Classical Fortran" book ,

may be a good one for start.
http://novian.web.ugm.ac.id/

Direct Link :
http://ul2.rapidshar.../0824708024.rar

[unique33]

If anyone need the book for print :
Follow this procedure in the Adobe Acrobat Professional :
1.
Adobe Acrobat Professional :
Menu File - Print ->
Printer : Name : Adobe PDF
click properties :
---------------------
in the layout tab select Portrait Orientation
choose Pages per sheet : 2
click advanced :
Graphic - Scaling 130%
---------------------
Now go to the Paper/Quality Tab
Color : Black & White
---------------------
Now go to the Adobe PDF Setting Tab :
Set Adobe PDF Page Size :A4 Or ....
---------------------
click OK
in the Print window :
Print Range : All or ....

Page Handling :
Page Scaling :None
Auto-Rotate and Center

Click OK :
Define a name for the new file and save it
you should wait here .
the Produced file will have aproximately 320 Mb size depending
on the dpi you have chosen in the Adobe Acrobat Properties.

-----------------------------------
2.
OPen the new file :
Menu File - Print ->
Printer : Name : your installed printer
click properties :
the scaling should be 100%

Print Range :
if you want to print on the back of your sheet :
Subset : First Odd
then Even Pages

Page scaling : None
Auto Rotate and Center

Click OK
finished.