nonlinear least square

Discussion in 'Python' started by Dmitry Rozmanov, Dec 23, 2003.

  1. Hi.

    I am looking for simple and convenient way to do nonlinear least square
    fit from within Python.

    I know that there are such a modules in Scientific Python and SciPy has
    it. But I would like to have it small and simple. Module with just this
    functionality. It can be not very fast, but in pure python for beter
    compatibility.

    Can you recommend such a module?

    Thank you in advance.

    Regards.

    ---Dmitry.
     
    Dmitry Rozmanov, Dec 23, 2003
    #1
    1. Advertising

  2. Dmitry Rozmanov

    John J. Lee Guest

    Dmitry Rozmanov <> writes:

    > I am looking for simple and convenient way to do nonlinear least
    > square fit from within Python.
    >
    > I know that there are such a modules in Scientific Python and SciPy
    > has it. But I would like to have it small and simple. Module with just
    > this functionality. It can be not very fast, but in pure python for
    > beter compatibility.

    [...]

    Konrad Hinsen's implementation in Scientific Python *is* in pure
    Python, and as a consequence is probably about as simple as you're
    going to get, I imagine. Is it dependencies on Scientific Python that
    you don't like? If so, just dive in and remove them, if it's that
    important to you -- I doubt you'll find a better pure Python
    implementation.

    I can certainly testify that SciPy's MINPACK wrapper works fine, too.
    I can dig out some example code if you like.


    John
     
    John J. Lee, Dec 23, 2003
    #2
    1. Advertising

  3. John J. Lee wrote:

    Thank you for the suggestion.

    Yes, I do not like to have to install Scientific Python just for this
    purpose. As I understand it has some modules which have to be compiled
    anyway.

    I would be really grateful for any example code you provide.

    BWT do you know such a standalone code for nonlinear equations set
    solving? Newton's method or something. I mean in Python only.

    Thank you again.

    Regards.

    ---Dmitry.

    > Konrad Hinsen's implementation in Scientific Python *is* in pure
    > Python, and as a consequence is probably about as simple as you're
    > going to get, I imagine. Is it dependencies on Scientific Python that
    > you don't like? If so, just dive in and remove them, if it's that
    > important to you -- I doubt you'll find a better pure Python
    > implementation.
    >
    > I can certainly testify that SciPy's MINPACK wrapper works fine, too.
    > I can dig out some example code if you like.
    >
    >
    > John
     
    Dmitry Rozmanov, Dec 23, 2003
    #3
  4. Dmitry Rozmanov

    William Park Guest

    Dmitry Rozmanov <> wrote:
    > John J. Lee wrote:
    >
    > Thank you for the suggestion.
    >
    > Yes, I do not like to have to install Scientific Python just for this
    > purpose. As I understand it has some modules which have to be compiled
    > anyway.
    >
    > I would be really grateful for any example code you provide.
    >
    > BWT do you know such a standalone code for nonlinear equations set
    > solving? Newton's method or something. I mean in Python only.


    Efficiency aside, it's easy to write them in Python. You can
    practically copy C-code line for line. Or, write algorithm from books.

    --
    William Park, Open Geometry Consulting, <>
    Linux solution for data management and processing.
     
    William Park, Dec 24, 2003
    #4
  5. Dmitry Rozmanov

    Chris Friedl Guest

    I recently found the Enthought python packages (2.2 & 2.3) which bundles a
    whole lot of scientific/numeric/visualisation python code (see
    www.enthought.com/python). Might be useful for you to get to the scipy
    stuff.

    Merry Xmas


    "Dmitry Rozmanov" <> wrote in message
    news:bsaj7i$26om$...
    > John J. Lee wrote:
    >
    > Thank you for the suggestion.
    >
    > Yes, I do not like to have to install Scientific Python just for this
    > purpose. As I understand it has some modules which have to be compiled
    > anyway.
    >
    > I would be really grateful for any example code you provide.
    >
    > BWT do you know such a standalone code for nonlinear equations set
    > solving? Newton's method or something. I mean in Python only.
    >
    > Thank you again.
    >
    > Regards.
    >
    > ---Dmitry.
    >
    > > Konrad Hinsen's implementation in Scientific Python *is* in pure
    > > Python, and as a consequence is probably about as simple as you're
    > > going to get, I imagine. Is it dependencies on Scientific Python that
    > > you don't like? If so, just dive in and remove them, if it's that
    > > important to you -- I doubt you'll find a better pure Python
    > > implementation.
    > >
    > > I can certainly testify that SciPy's MINPACK wrapper works fine, too.
    > > I can dig out some example code if you like.
    > >
    > >
    > > John

    >
     
    Chris Friedl, Dec 24, 2003
    #5
  6. Thank you for the link. It is of use for windows installation definitely.

    Regards.

    ---Dmitry.

    Chris Friedl wrote:
    > I recently found the Enthought python packages (2.2 & 2.3) which bundles a
    > whole lot of scientific/numeric/visualisation python code (see
    > www.enthought.com/python). Might be useful for you to get to the scipy
    > stuff.
    >
    > Merry Xmas
    >
    >
    > "Dmitry Rozmanov" <> wrote in message
    > news:bsaj7i$26om$...
    >
    >>John J. Lee wrote:
    >>
    >>Thank you for the suggestion.
    >>
    >>Yes, I do not like to have to install Scientific Python just for this
    >>purpose. As I understand it has some modules which have to be compiled
    >>anyway.
    >>
    >>I would be really grateful for any example code you provide.
    >>
    >>BWT do you know such a standalone code for nonlinear equations set
    >>solving? Newton's method or something. I mean in Python only.
    >>
    >>Thank you again.
    >>
    >>Regards.
    >>
    >>---Dmitry.
    >>
    >>
    >>>Konrad Hinsen's implementation in Scientific Python *is* in pure
    >>>Python, and as a consequence is probably about as simple as you're
    >>>going to get, I imagine. Is it dependencies on Scientific Python that
    >>>you don't like? If so, just dive in and remove them, if it's that
    >>>important to you -- I doubt you'll find a better pure Python
    >>>implementation.
    >>>
    >>>I can certainly testify that SciPy's MINPACK wrapper works fine, too.
    >>>I can dig out some example code if you like.
    >>>
    >>>
    >>>John

    >>

    >
    >
     
    Dmitry Rozmanov, Dec 24, 2003
    #6
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Replies:
    0
    Views:
    1,255
  2. Replies:
    1
    Views:
    326
  3. Replies:
    2
    Views:
    520
  4. Uwe Kotyczka

    Nonlinear least square problem

    Uwe Kotyczka, Apr 3, 2008, in forum: Python
    Replies:
    2
    Views:
    378
    Michael Press
    Apr 5, 2008
  5. AAaron123
    Replies:
    0
    Views:
    618
    AAaron123
    Oct 3, 2008
Loading...

Share This Page