Rolling multidimensional function in pandas

Question

Let's say, I have the following code.

import numpy as np
import pandas as pd

x = pd.DataFrame(np.random.randn(100, 3)).rolling(window=10, center=True).cov()

For each index, I have a 3x3 matrix. I would like to calculate eigenvalues and then some function of those eigenvalues. Or, perhaps, I might want to compute some function of eigenvalues and eigenvectors. The point is that if I take x.loc[0] then I have no problem to compute anything from that matrix. How do I do it in a rolling fashion for all matrices?

Thanks!

Sumanth · Accepted Answer · 2019-01-21 05:09:02Z

1

You can use the analogous eigenvector/eigenvalue methods in spicy.sparse.linalg.

import numpy as np
import pandas as pd
from scipy import linalg as LA

x = pd.DataFrame(np.random.randn(100, 3)).rolling(window=10, center=True).cov()
for i in range(len(x)):
    try:
        e_vals,e_vec = LA.eig(x.loc[i])
        print(e_vals,e_vec)
    except:
        continue

If there are no NaN values present then you need not use the try and except instead go for only for loop.

edited Jan 21, 2019 at 5:09

answered Jan 20, 2019 at 5:48

Sumanth

5075 silver badges13 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

freevillage Over a year ago

What I don't understand is why I can calculate a covariance matrix for each "window x 3" window but I cannot calculate a more general function. Of course, I can do another loop after that but it seems wasteful. Why go over the matrix twice?

Collectives™ on Stack Overflow

Rolling multidimensional function in pandas

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related