Back to Community
Encountering troubles using np.corrcoef as a CustomFactor in Research

Hello All,
I am attempting to create a custom factor tear sheet from Alpha 6 of the 101Alphas paper. Link to the paper is here: https://goo.gl/jPJwg8. I have ruthlessly copied the Factor Tear Sheet notebook from the Lecture series. Below I create my factor, which according to the strange documentation that is 101Alphas is: (-1 * correlation(open, volume, 10)). Basically I need to take a ten day correlation factor between the open price and volume and present that in its negative form as my factor. Now to the problem at hand:

I am confident it is a python array issue, but everytime I attempt to pass the open and volume values to corrcoef and get the 0-1 matrix value from it, I always get this "zero-size array to reduction operation minimum which has no identity" error. Please help me debug this issue.

Thanks

Loading notebook preview...
Notebook previews are currently unavailable.