Usage for linfit

Return to directory of Todd's CASA extensions

Basic linear function fitter with error bars in the y-axis data points. Uses scipy.optimize.leastsq(). Accepts either lists or arrays. See also spectralindex. For linear fits with uncertainties in both axes, see orthogonalDistanceRegression.


lf = au.linfit()
lf.linfit(x, y, yerror, pinit=[0.57, 3.4], plot=False, plotfile=None, xlabel=None, 
     ylabel=None, title=None, residual=False, excludeXrange=[], returnResiduals=False) 
  • Input:
    • x, y: x and y axis values
    • yerror: uncertainty in the y-axis values
    • pinit contains the initial guess of [slope, intercept]
    • residual: if True, show the residual of the fit in second panel
    • excludeXrange: exclude a range of x-axis values when doing the fit e.g. [109,116] to avoid 109-116 GHz points
    • returnResiduals: if True, return [slope,intercept,data-fit]
  • Output:
    • The fit result as: [slope, y-intercept], or [slope, y-intercept, data-fit]


CASA<5> lf = au.linfit()
CASA<6> x = [1,2,3,4,5]
CASA<7> y = [.43,.96,1.24,1.67,1.86]
CASA<8> yerror = [0.1,0.1,0.1,0.2,0.2]
CASA<9> lf.linfit(x, y, yerror, xlabel='my x label', ylabel='my y label', title='test', plotfile='test.png', plot=True)
  Out[9]: array([ 0.31652522,  0.30612593])

-- ToddHunter - 2012-09-10
  • test.png:
Topic revision: r5 - 2014-09-02, ToddHunter
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding NRAO Public Wiki? Send feedback