BUG: ewm*() interpretation of min_periods is off by one

[seth-p]

Closes #7884.

This addresses only the ewm*() functions' off-by-one interpretation of min_periods.
It doesn't address the inconsistency between the ewm*() functions and the rolling_*() functions in the meaning of min_periods.

[jreback]

are there test for min_periods 1 and 0?
if not can u add pls

thxs

[seth-p]

Just removing gratuitous "a" from function names, and adding ewmvol.

[seth-p]

OK, will add tests for min_periods 0 and 1.

[seth-p]

In the process of adding tests for min_periods = 0, 1, I noticed the following, which is really messing up the testing. I think all these functions should return NaN for a single value.

I just noticed that ewmvar, ewmstd, ewmvol, ewmcov, rolling_var, rolling_std, returns 0.0 for a single value (assuming min_periods=0); whereas Series.std, Series.var, ewmcorr, expanding_cov, expanding_corr, expanding_std, expanding_vol, and expanding_var, rolling_cov, and rolling_corr all return NaN for a single value.

[seth-p]

I've updated the code to add tests for min_periods = 0, 1, 2, leaving the single-value behavior noted above unchanged. If want to, can address that in a separate issue/PR: #7900.

[jreback]

add the issue at the end

[seth-p]

The Travis CI build passed.

[jreback]

change to issue and repush and lmk (don't need to wait for green)

[seth-p]

D'oh! Am away from computer now; will do later.

On Aug 1, 2014, at 6:40 PM, jreback [email protected] wrote:

In doc/source/v0.15.0.txt:

@@ -80,6 +80,11 @@ API changes
ewma(Series([1., None, 100.]), com=2.5, ignore_na=True) # pre-0.15.0 behavior
ewma(Series([1., None, 100.]), com=2.5, ignore_na=False) # default

+- :func:ewma, :func:ewmstd, :func:ewmvol, :func:ewmvar, :func:ewmcorr, and :func:ewmcov

• now set to NaN the first min_periods-1 entries of the result (for min_periods>1).
• Previously the first min_periods entries of the result were set to NaN.
• The new behavior accords with the existing documentation. (:issues:7884)
  change to issue and repush and lmk (don't need to wait for green)

—
Reply to this email directly or view it on GitHub.

[seth-p]

I re-pushed an update removing the spurious s.

[jreback]

thanks!