Freeware to do "curve fitting"

[Wishmaster]

Hello!

Do you know any freeware program to do curve fitting? I currently use an
ancient "never-expiring" shareware program that doesn't work correctly on
Windows XP.

I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a set of points
(x,y).

 

[Jeff Needle]

I assume you're talking about finding a line of regression? Don't most
spreadsheet programs, like Excel, offer the a and b coordinates of the
regression line? Perhaps you need something more elaborate. And I may be
wrong about Excel doing this.

In the meantime, check this out:

http://www.rjsweb.fsnet.co.uk/graph/

If it works as it says it does, would you mind letting me know? I need
such a program, but am waiting to hear if this one is as good as it looks.

Thanks.

 

[Jim Byrd]

Hi Wishmaster - IMO, DataFit by Oakdale Engineering
(http://www.oakdaleengr.com/datafit.htm) is the best around - I've used it
for many years, but very definitely not free. Here's one that is, but I
haven't used it, and therefore can't recommend either for or against:

http://crmht.cnrs-orleans.fr/pot/software/focus.html It appears to be
oriented to spectroscopic data, but might be of use.

Some of those listed here are free also and may fit your needs: (See
CurvFit and RJS Graph, for example)
http://dmoz.org/Science/Math/Software/Graphing/

 

[Terry Russell]

Hello!

Do you know any freeware program to do curve fitting? I currently use an
ancient "never-expiring" shareware program that doesn't work correctly on
Windows XP.
I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a set of points
(x,y).

google spline freeware

depends whether you want best fit line or a
curve that includes all points


A lot of software collapses with faster machines.
Old Borland Turbo pascal had a bug that fails over
(X?) mhz.

google borland turbo pascal crt bug

http://www.merlyn.demon.co.uk/pas-r200.htm
for one
ftp://ftp.heise.de/pub/ct/ctsi/ctbppat.zip
maybe you can patch the old software and avoid a learning curve

 

[wald]

Do you know any freeware program to do curve fitting? I
currently use an ancient "never-expiring" shareware program that
doesn't work correctly on Windows XP.

I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a
set of points
(x,y).

Probably very much overkill for your purposes, but R is an open
source software package for statistical computing/modelling. You can
use it to fit *any* function to a set of data points, as long as you
are willing to first learn a bit of R programming. The language is
actually S, so if you've ever worked with Splus or alike you'll
already be familiar...

http://r-project.org/

Regards,
Wald

 

[old jon]

Hello!

Do you know any freeware program to do curve fitting? I currently use an
ancient "never-expiring" shareware program that doesn't work correctly on
Windows XP.

I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a set of points
(x,y).

Have you tried making your existing program Compatible to see if it will
work ?
Rt click program\Properties\ Compatibility. Gotta be worth a try. BW..OJ

 

[REM]

"Wishmaster" <[email protected]> wrote:

Do you know any freeware program to do curve fitting? I currently use an
ancient "never-expiring" shareware program that doesn't work correctly on
Windows XP.

I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a set of points
(x,y).

KyPlot?

http://www.woundedmoon.org/win32/kyplot.html


I'm not certain that this does what you need. Check it out though.
It's a high quality graphing and statistical program that went
commercial.

Or:

http://jansfreeware.com/jftools.htm

"EasyGrapher

4-August-2002 size: 339kb
Plot graphs of one or more series of functions or datapoints.
Features: Store/Load graphs in XML format; save graph as scalable EMF
image or web ready PNG image; Polynomial curve fitting; zoom and pan."


A couple of freeware utilities that might be of interest:

http://www.xray.cz/ecm-cd/soft/win/index0065.html

 

[Dick Kistler]

Probably very much overkill for your purposes, but R is an open
source software package for statistical computing/modelling. You can
use it to fit *any* function to a set of data points, as long as you
are willing to first learn a bit of R programming. The language is
actually S, so if you've ever worked with Splus or alike you'll
already be familiar...

http://r-project.org/

Dead! How about http://www.r-project.org/

Dick Kistler

 

[Dick Kistler]

Dead! How about http://www.r-project.org/

Dick Kistler

Later, try http://members.aol.com/johnp71/javasta2.html for many
possibilities.

Dick

 

[ellis_jay]

Hello!

Do you know any freeware program to do curve fitting? I currently use
an ancient "never-expiring" shareware program that doesn't work
correctly on Windows XP.

I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a set of
points (x,y).

Look around here:

http://www.merlot.org/Home.po

http://faculty.vassar.edu/lowry/VassarStats.html




--

Their ethics are a short summary of police ordinances: for them the
most important thing is to be a useful member of the state, and to air
their opinions in the club of an evening; they have never felt the
homesickness for something unknown and far away, nor the depths which
consists in being nothing at all. ___________Soren Kierkegaard

Ellis_jay

 

[Wald]

...

"Dick Kistler" wrote...

Sorry 'bout that. I got confused because of the auto-URL searching
Firefox does behind the screens. IE doesn't find the page, you're right.

Later, try http://members.aol.com/johnp71/javasta2.html for many
possibilities.

True. However, if the OP is looking for a full-featured statistical
package in the spirit of SAS/STAT, then R is the best choice. Active
development, contributions by the cream of statisticians around the
planet, has the latest statistical techniques.

Regards,
Wald

 

[Wishmaster]

In the meantime, check this out:

If it works as it says it does, would you mind letting me know? I need
such a program, but am waiting to hear if this one is as good as it looks.

It works, but it's almost useless, due to this:

- Very slow here (a 3.0GHz Pentium IV)
- the program thinks that it is MS Office and
- horrible interface (vertical tabs! ouch! I felt like using the utilities
from my GeForce FX5200 video card to rotate the video ).

I ended uninstalling it.

 

[ggrothendieck]

Do you know any freeware program to do curve fitting? [...]
I don't know how to explain exactly, but:
curve fitting is like trying to discover an equation from a set of points
(x,y).

There are two very good nonlinear regression programs, curveExpert
and nlreg. They are shareware rather than freeware but you can try
them out for free. The first is used for fitting a set of x,y points
to a large number of possible curves rapdily. The second has an
entire language built into it yet it is relatively easy to
use and is quite powerful. Both have builtin graphics.

I typically use R for this, which was mentioned here, is free and
very powerful, but it does have a substantial learning curve so for
certain focused tasks that you want to accomplish quickly the two
programs mentioned here would be easier to use.

 

[Wishmaster]

Wishmaster wrote:

I typically use R for this, which was mentioned here, is free and
very powerful, but it does have a substantial learning curve so for
certain focused tasks that you want to accomplish quickly the two
programs mentioned here would be easier to use.

I will try learning R, seems like a nice software. Any tutorials?

[]s

 

[ggrothendieck]

I will try learning R, seems like a nice software. Any tutorials?

Enter the single letter R into google to get to the R home page.

1. On the home page in the left hand pane, under Documentation
click on Manuals and choose the Introduction to R manual.

2. Also on the home page under Documentation, if you click on Other
you will get a list of intro and other documents written
by others.

3. From #2 you can click on 'contributed documentation'
to get even more. For example, the R Reference Card by Tom Short
on this page is quite helpful.

You can also get to much of this right from the Help menu
from within R on Windows.

 

[ggrothendieck]

I will try learning R, seems like a nice software. Any tutorials?

Enter the single letter R into google to get to the R home page.

1. On the home page in the left hand pane, under Documentation
click on Manuals and choose the Introduction to R manual.

2. Also on the home page under Documentation, if you click on Other
you will get a list of intro and other documents written
by others.

3. From #2 you can click on 'contributed documentation'
to get even more. For example, the R Reference Card by Tom Short
on this page is quite helpful.

You can also get to much of this right from the Help menu
from within R on Windows.