[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
## Question about use of gsl_multifit_linear_est

**From**: |
mhkelley2017 |

**Subject**: |
Question about use of gsl_multifit_linear_est |

**Date**: |
Wed, 15 Jul 2020 17:38:51 -0600 |

I am confused about the use of gsl_multifit_linear capabilities. I already
successfully use gsl_fit_linear. I can successfully used
gsl_multifit_linear to actually extract the parameters to fit a polynomial
model to a dataset. My confusion is how to extract the error estimates of
the overall fit. I know (in principle) how to use the covariance matrix to
calculate the error estimate myself, but wanted to just use
gsl_multifit_linear_est, Which is provided for that purpose.
Unfortunately, I can't make sense of the manual page.
int gsl_multifit_linear_est(const
<https://www.gnu.org/software/gsl/doc/html/vectors.html#c.gsl_vector>
gsl_vector * x, const
<https://www.gnu.org/software/gsl/doc/html/vectors.html#c.gsl_vector>
gsl_vector * c, const
<https://www.gnu.org/software/gsl/doc/html/vectors.html#c.gsl_matrix>
gsl_matrix * cov, double * y, double * y_err)
This function uses the best-fit multilinear regression coefficients c and
their covariance matrix cov to compute the fitted function value y and its
standard deviation y_err for the model at the point x.
This implies that I should provide some value for "x" (along with the fit
coefficients and covariance matrix) to generate the corresponding values for
"y" and "y_err". But the function wants me to provide a gsl_vector of x
values. Why shouldn't that return a gsl_vector of y and y_err values?
I'm confused - anyone able to clarify?
Thanks in advance.

[Prev in Thread] |
**Current Thread** |
[Next in Thread] |

**Question about use of gsl_multifit_linear_est**,
*mhkelley2017* **<=**