Winsorized Mean and Standard Deviation

[Wheeliam]

Hi all,

I have a slight problem. I need to winsorize the means and standard deviations in my code. The code is a modification of what I have bumped into online (with a few add-ons and corrections).

The mean is denoted as:

e = numpy.mean( r )
ps. I have added spaces between brackets and r - cos it was changing it to r in circle

How would I go about winosrizing it? Here is an article on how to deal with this issue, although given how new I am to Python, I would need someone's guidance in implementation.

https://docs.scipy.org/doc/scipy/referen...orize.html

The only other instance of the "mean" I have in my code is when I calculate Information Ratio:

def information_ratio(returns, benchm):
diff = returns - benchmark
return numpy.[b]mean[/b](diff) / vol(diff)

Furthermore, would the same methodology (as for the winosrized mean) apply to the standard deviation (winsorizing it)?

Again, the only instance of the standard deviation I have listed in my code is:

return numpy.std(returns)

Please advise.

Thank you