Numpy polyfit - covariance matrix

Question

When fitting a straight line to a set of data, weighted with errors, I was expecting polyfit to return a 2x2 covariance matrix from which I could square root the diagonal elements to find the uncertainty in the coefficients, but I don't.

Here's a minimum working example:

from numpy import polyfit

x=[1.,2.,3.,4.,5.,6.,7.,8.,9.]
y=[1.,3.,2.,4.,5.,6.,6.,8.,9.]

yerr=[10.,5.,3.,2.,10.,10.,10.,10.,10.]

linear = polyfit(x,y, 1, w=yerr, full=True)

print(linear)

With output:

(array([ 0.95730623,  0.11546722]), array([ 114.79556527]), 2, array([ 1.38182992,  0.30090875]), 1.9984014443252818e-15)

Thanks!

Answer 1

The document says:

• cov : bool, optional

  Return the estimate and the covariance matrix of the estimate If full is True, then cov is not returned.

So set cov=True and full=False(default setting) will return the convariance matrix.