Calculating weighted moving average using pandas Rolling method

Question

I calculate simple moving average:

def sma(data_frame, length=15):
    # TODO: Be sure about default values of length.
    smas = data_frame.Close.rolling(window=length, center=False).mean()
    return smas

Using the rolling function is it possible to calculate weighted moving average? As I read in the documentation, I think that I have to pass win_type parameter. But I'm not sure which one I have to choose.

Here is a definition for weighted moving average.

Thanks in advance,

Have a look at np.average which supports weights to be provided. — Nickil Maveli
I would file an issue on Github. The documentation is indeed misleading. It suggests one could pass custom weights, but it doesn't explain how one would do so, and there are no examples in the repository where this is done. — dangom

Dthal Dthal · Accepted Answer · 2016-09-28T09:57:29

Yeah, that part of pandas really isn't very well documented. I think you might have to use rolling.apply() if you aren't using one of the standard window types. I poked at it and got this to work:

>>> import numpy as np
>>> import pandas as pd
>>> d = pd.DataFrame({'a':range(10), 'b':np.random.random(size=10)})
>>> d.b = d.b.round(2)
>>> d
   a     b
0  0  0.28
1  1  0.70
2  2  0.28
3  3  0.99
4  4  0.72
5  5  0.43
6  6  0.71
7  7  0.75
8  8  0.61
9  9  0.14
>>> wts = np.array([-1, 2])
>>> def f(w):                        
        def g(x):
            return (w*x).mean()
        return g
>>> d.rolling(window=2).apply(f(wts))
     a      b
0  NaN    NaN
1  1.0  0.560
2  1.5 -0.070
3  2.0  0.850
4  2.5  0.225
5  3.0  0.070
6  3.5  0.495 
7  4.0  0.395
8  4.5  0.235
9  5.0 -0.165

I think that is correct. The reason for the closure there is that the signature for rolling.apply is rolling.apply(func, *args, **kwargs), so the weights get tuple-unpacked if you just send them to the function directly, unless you send them as a 1-tuple (wts,), but that's weird.

Calculating weighted moving average using pandas Rolling method

1 Answers