A GMT version of linearfitxy

[joa-quim]

Hi Rafael,

I've ported your linearfitxy function to work in GMT.jl and in the migration process the Distributions and Plots dependencies were dropped (replaced by GMT.jl functions alone).

Taking one of your examples I can now do

D = Misc.linearfitxy([0.0, 0.9, 1.8, 2.6, 3.3, 4.4, 5.2, 6.1, 6.5, 7.4], [5.9, 5.4, 4.4, 4.6, 3.5, 3.7, 2.8, 2.8, 2.4, 1.5], sx=1 ./ sqrt.([1000., 1000, 500, 800, 200, 80,  60, 20, 1.8, 1]), sy=1 ./ sqrt.([1., 1.8, 4, 8, 20, 20, 70, 70, 100, 500]));

>>> [± σ]  Y = (5.4799 ± 0.3592) + (-0.4805 ± 0.0706)*X
>>> [± 95% CI]  Y = (5.4799 ± 0.8284) + (-0.4805 ± 0.1629)*X
>>> Pearson ρ = -0.943;  Goodness of fit = 1.218

# This one was entirely re-written in *GMT.jl* and allows tuning all plotting parameters
plot_linfitxy(D, band_ab=true, band_CI=true, ellipses=true, show=true)

and get the figure bellow.

What I would like to ask you is if you would like to make a PR to GMT.jl with the contents of the ported function (in the zip attached to this post) so that the credits for this work are more openly visible.

Cheers

Joaquim

linearfitxy.zip

[rafael-guerra-www]

Hi Joaquim,

Thanks for your message and interest. And I am happy to see that everything (and more) can be done in GMT. I will be glad to make the PR to GMT, with a disclaimer (as per the package notes) :)

Regards, Rafael