How to fix a parameter when using curve_fit ?

[Spark Liang]

Hi, would someone be so kind to tell me how to fix some parameters when using curve_fit ?
I googled it and found that I may use scipy.odr or mpfit, but they seem rather complicated.
I also searched the maillist, someone said it can be done by by writing a nested function or a class. but how to do it?

_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user

[josef.pktd]

On Fri, Jun 10, 2011 at 8:01 AM, Spark Liang <[hidden email]> wrote:
> Hi, would someone be so kind to tell me how to fix some parameters when
> using curve_fit ?
> I googled it and found that I may use scipy.odr or mpfit, but they seem
> rather complicated.
> I also searched the maillist, someone said it can be done by by writing a
> nested function or a class. but how to do it?

a full version is at http://docs.python.org/library/functools.html in
the example for functools.partial

I usually prefer to use a class, where you set the fixed parameters in
the __init__ and access it with self.a(ttributename) inside the
function.

Josef

>
> _______________________________________________
> SciPy-User mailing list
> [hidden email]
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>
_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user

[x.piter]

Hi.
In matlab I was doing it using global variables, but is has to be
better way with nested functions.
The idea is to rewrite your fitted function in the way that only not
fixed parameters  are fed to optimization routine,
but fixed variables are still available to it.
Hope it helps a bit.
Best.
Petro.





On Fri, Jun 10, 2011 at 2:01 PM, Spark Liang <[hidden email]> wrote:

> Hi, would someone be so kind to tell me how to fix some parameters when
> using curve_fit ?
> I googled it and found that I may use scipy.odr or mpfit, but they seem
> rather complicated.
> I also searched the maillist, someone said it can be done by by writing a
> nested function or a class. but how to do it?
>
> _______________________________________________
> SciPy-User mailing list
> [hidden email]
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>

_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user

[Spark Liang]

thanks, Petro.
I remember that matlab has such options in optimize toolsbox

On Fri, Jun 10, 2011 at 8:53 PM, Piter_ <[hidden email]> wrote:

Hi.
In matlab I was doing it using global variables, but is has to be
better way with nested functions.
The idea is to rewrite your fitted function in the way that only not
fixed parameters  are fed to optimization routine,
but fixed variables are still available to it.
Hope it helps a bit.
Best.
Petro.

On Fri, Jun 10, 2011 at 2:01 PM, Spark Liang <[hidden email]> wrote:

> Hi, would someone be so kind to tell me how to fix some parameters when
> using curve_fit ?
> I googled it and found that I may use scipy.odr or mpfit, but they seem
> rather complicated.
> I also searched the maillist, someone said it can be done by by writing a
> nested function or a class. but how to do it?
>

> _______________________________________________
> SciPy-User mailing list
> [hidden email]
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>
_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user

_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user

[Spark Liang]

thanks, Josef.

I have done it by using mpfit, though it's not very easy.

On Fri, Jun 10, 2011 at 8:35 PM, <[hidden email]> wrote:

On Fri, Jun 10, 2011 at 8:01 AM, Spark Liang <[hidden email]> wrote:
> Hi, would someone be so kind to tell me how to fix some parameters when
> using curve_fit ?
> I googled it and found that I may use scipy.odr or mpfit, but they seem
> rather complicated.
> I also searched the maillist, someone said it can be done by by writing a
> nested function or a class. but how to do it?

a full version is at http://docs.python.org/library/functools.html in
the example for functools.partial

I usually prefer to use a class, where you set the fixed parameters in
the __init__ and access it with self.a(ttributename) inside the
function.

Josef

>
> _______________________________________________
> SciPy-User mailing list
> [hidden email]
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>
_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user

_______________________________________________
SciPy-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/scipy-user