Yes, thanks Giovanni. Numpy worked.
On Jul 21, 1:51 am, "giovanni.marche...@ds.unifi.it"
wrote:
> I read in the listing of the function
>
> warnings.warn("""\
> scipy.stats.corrcoef is deprecated; please update your code to use
> numpy.corrcoef.
> Please note that:
> - numpy.corrcoef ro
I read in the listing of the function
warnings.warn("""\
scipy.stats.corrcoef is deprecated; please update your code to use
numpy.corrcoef.
Please note that:
- numpy.corrcoef rowvar argument defaults to true, not false
- numpy.corrcoef bias argument defaults to false, not true
""", De
The bias defaults to False, which means the covariance is calculated
with a division by N-1 instead of N. This is what should be used
when sampling from a population. If you have data on an entire
population, you want the biased estimator which divides by N. For
large N this will make very lit
